Why does the figure change so drastically? And how can I solve it?
What are you referring yo specifically? The data plots seem to be identical.
Sidenote: there seems to be a bug in the plot viewer, as the axis are a bit chaotic..
Do you mean the x/y intersection?
Does it save the code in the uncommitted changes?
Did you run the agent with the --docker tag?
Hi @<1639799308809146368:profile|TritePigeon86> , what is the use case for passing multiple callbacks? Why not have it in the same function simply?
RoughTiger69 , can you provide the configuration of the autoscaler please?
AlertCrow40 , by the way. ClearML already has an integrated tool to work on a jupyter notebook.
In a couple of lines it will open a jupyter notebook for you to work with. Further reading here: https://clear.ml/docs/latest/docs/apps/clearml_session/
🙂
Hi JitteryCoyote63 , I don't believe this is possible. Might want to open a GitHub feature request for this.
I'm curious, what is the use case? Why not use some default python docker image as default on agent level and then when you need a specific image put into the experiment configuration?
You're still using both n1-standard-1 and nvidia/cuda:10.2-runtime-ubuntu18.04
Hi SmugSnake6 , can you please elaborate on what exactly is happening and what you were expecting to happen?
Hi ReassuredArcticwolf33 , what are you trying to do and how is it being done via code?
TimelyPenguin76 , MammothGoat53 , I think you shouldn't call Task.init() more than once inside a script
How did you add the parameters to the pipeline? Did you refer to this example?
None
The way that community server works, yes. All your experiments are connected to a specific workspace/user
Hi GaudyPig83 , when is this happening?
Do you have a code snippet that reproduces this?
Please run the following commands and share the results. Chances are that somehow the default mappings that we apply on the index creation were not applied to your events scalar index.
- First run the following command
curl -XGET "localhost:9200/_cat/indices/events-training_stats_scalar-*"
- And then for each of the returned indices run the following:
curl -XGET "localhost:9200/<index_name>/_mappings"
I don't think you can currently assign cpu cores to the agents. They just use the resources they have in cpu mode
I would suggest adding print outs during the code to better understand when this happens
Hi ShallowGoldfish8 , can you elaborate please? You mean train with different data?
ScaryLeopard77 , Hi! Is there a specific reason to the aversion from pipelines? What is the use case?
"continue with this already created pipeline and add the currently run task to it"
I'm not sure I understand, can you please elaborate? (I'm pretty sure it's a pipelines feature)
BattyLizard6 Hi!
In the basic scenario where the database is empty on the new server you'd simply need to copy /opt/clearml/ from the old server to the new. In the case that a data merge is needed I'm not sure, let me check 🙂
You mean the you want a custom x axis name?
Hi @<1603198163143888896:profile|LonelyKangaroo55> , you certainly can. I think you need to enable editing these configurations but it certainly is possible with some tinkering 🙂
I'm afraid you can't do that. Each user has his own workspace and users join to that workspace. If you would like to have a new 'owner' you would need to add your users to that new person's workspace
I've never worked with JupyterHub and have little experience with notebooks. What does it do in relation to notebooks?
Hi @<1587615463670550528:profile|DepravedDolphin12> , your cache is defined by your clearml.conf you can see where it points and delete that folder 🙂
Check the environment variables, maybe test with export maybe there's some env var hiding there 🙂
Usage quote is calculated a few times a day. The new stats should be reflected in a few hours
HelplessCrocodile8 Hi!
What do you mean exactly by "drive mapping to save..." ? are you referring to upload artifacts ? are those models ? data?
Hi ShallowGoldfish8 , what versions of ClearML & ClearML-Agent are you using?