Reputation
Badges 1
25 × Eureka!. However, despite having imported the required types from theΒ
typing
Β library in the script where the function decorated withΒ
PipelineDecorator.component
Β is defined, later in the generated script theΒ
typing
Β library is not imported outside the scope of the function
Actually the typing part is not passed to the "created step" , because there are no global imports, for eexample:
` def step(a: pd.DataFrame):
import pandas as pd
...
ReassuredTiger98 can you send the full log?
Also, what's the clearml-agent version?
fyi: we fixed an issue where the default order of the conda repositories cause pytorch to be installed form the conda forge instead of the pytorch repo, making it the cpu version instead of the gpu version:
This is the correct conda repo orderL
https://github.com/allegroai/clearml-agent/blob/cb6bdece39751eaef975287609b8bab603f116e5/docs/clearml.conf#L66
or at least stick to the requirements.txt file rather than the actual environment
You can also for it to log the requirements.txt withTask.force_requirements_env_freeze(requirements_file="requirements.txt") task = Task.init(...)
Hi @<1523701868901961728:profile|ReassuredTiger98>
Could you send the full log ? Also what's the clearml-agent
version?
assuming you have http://hparams.my _param
my suggestion is:
` @hydra.main(config_path="solver/config", config_name="config")
def train(hparams: DictConfig):
task = Task.init(hparams.task_name, hparams.tag)
overrides = {'my_param': hparams.value}
task.connect(overrides, name='overrides')
in remote this will print the value we put in "overrides/my_param"
print(overrides['my_param'])
now we actually use overrides['my_param'] `Make sense ?
I mean to reduce the API calls without reducing the scalars that are logged, e.g. by sending less frequent batched updates.
Understood,
In my current trials I am using up the API calls very quickly though.
Why would that happen?
The logging is already batched (meaning 1API for a bunch of stuff)
Could it be lots of console lines?
BTW you can set the flush period to 30 sec, which would automatically collectt and batch API calls
https://github.com/allegroai/clearml/blob/25df5efe7...
save off the "best" model instead of the last
Should be relatively easy to update on the main Task the model with the best performance, no?
Hi GloriousPenguin2
Had to do some linux updates and redeploy clearml server, now i can access web UI & the service only if i do port-forwarding to that remote machine
So you are saying before you were able to directly browse to the server, but now you need a "jump box" ?
Hi BoredSquirrel45
as of today, my required packages aren't being recognized in cloned
Are you saying you are editing the code directly in the cloned Task, then enqueue the Task an the agent does not "auto recognize" the package ?
-- I've been running my script from VSCode for the first time,
In the initial Task (the one created when running inside VSCode) do you have all the packages listed in the "Installed Packages" section ?
Wow, thank you very much. And how would I bind my code to task?
you mean the code that creates pipeline Tasks ?
(remember the pipeline itself is a Task in the system, basically if your pipeline code is a single script it will pack the entire thing )
ReassuredTiger98 are you saying you want to be able to run the pipeline as a standalone and as "remote pipeline",
Or is this for a specific step in the pipeline that you want to be able to run standalone/pipelined ?
(I suspect you are correct, but I'm missing some information in order to understand where the problem is)
WackyRabbit7 can you send mock code that explains how you create the pipeline ?
Just making sure, after the pipe
object is created, you can call Task.current_task() , is that correct?
This doesn't seem to be running inside a container...
What's the clearml-agent launch command you are using ? (i.e. do you have --docker flag)
Okay now let's try the final lines:$LOCAL_PYTHON -m virtualenv /root/venv /root/venv/bin/python3 -m pip install git+
Nooooooooooooooooooooooo
Yes please, just to verify my hunch.
I think that somehow the docker mounts the agent is creating are (for some reason) messing it up.
Basically you can just run the following (it will do everything automatically) (replace the <TASK_ID_HERE> with the actual one)
` docker run -it --gpus "device=1" -e CLEARML_WORKER_ID=Gandalf:gpu1 -e CLEARML_DOCKER_IMAGE=nvidia/cuda:11.4.0-devel-ubuntu18.04 -v /home/dwhitena/.git-credentials:/root/.git-credentials -v /home/dwhitena/.gitconfig:/root/.gitconfig ...
Also in the same open docker session, can you try:$LOCAL_PYTHON -m clearml_agent execute --disable-monitoring --id <task_id_here>
Where the Task ID is one of the failed executions (only reset it before)
1e876021bbef49a291d66ac9a2270705
just make sure you reset it π
Can you send the console output of this entire session please ?
So what is the difference?!
Yes my bad π
Let's try again:
` docker run -it --gpus "device=1" -e CLEARML_WORKER_ID=Gandalf:gpu1 -e CLEARML_DOCKER_IMAGE=nvidia/cuda:11.4.0-devel-ubuntu18.04 -v /home/dwhitena/.git-credentials:/root/.git-credentials -v /home/dwhitena/.gitconfig:/root/.gitconfig -v /tmp/.clearml_agent.7rjdh80a.cfg:/root/clearml.conf -v /tmp/clearml_agent.ssh.ppsd9sze:/root/.ssh -v /home/dwhitena/.clearml/apt-cache.1:/var/cache/apt/archives -v /home/dwhitena/.clearml/pip-cache:/root/.cache/pip ...
Hi SmallDeer34
Is the Dataset in clearml-data ? If it is then Dataset.get().get_local_copy() will get you a cached local copy of the entire dataset.
If it is not, then you can use StorageManager.get_local_copy(url_here) to download the dataset.
- Any Argparser is automatically logged (and later can be overridden from the UI). Specifically HfArgumentParser will be automatically logged https://github.com/huggingface/transformers/blob/e43e11260ff3c0a1b3cb0f4f39782d71a51c0191/examples/pytorc...
Hi SubstantialElk6
Generally speaking here, the idea is that actual code creates a Dataset (i.e. Dataset class created from code), plus you can add some metric reporting (like table reporting) to create a preview of the data stored for better visibility, or maybe create some statistics as part of the data ingest script. Then this ingest code can be relaunched / automated. The created Dataset itself can be tagged renamed added key/value for better cataloging. wdyt?
SubstantialElk6 could you add a github issue to set the direct url for the vscode as a parameter to the cleaml-session?
We already have --vscode-version
we could either extend it to include a direct url, or add a new argument.
wdyt ?
But from the log it seems that:
you are not running as root in the docker? Python3.8 is installed (and not python 3.6 as before)
in the UI, find the task (just search for the Task ID, it will find it), then tight click it, and select "reset"
Hi MiniatureCrocodile39
Which packages to you need to run the viewer? I suppose dicom reader is a must?