I see now.
Let's assume you know which snapshot that was:
` prev_task = Task.get_task(task_id='the_first_training_task_id')

get the second from last checkpoint

task.models['output'][-2].url
prev_scalars = prev_task.get_reported_scalars()
new_task = Task.init('example', 'new task')
logger = new_task.get_logger()

do some fpr loop and report the prev_scalars with logger.report_scalars

new_task.flush(wait_for_uploads=True)
new_task.set_initial_iteration(22000)

start the train `

Basically try with the latest RC 🙂
pip install trains 0.15.2rc0

SubstantialKoala71 not sure I follow, what's the goal here ?

We should update the readme 🙂

is the model overridden or its version is automatically increased?

You will have another model, with the same name (assuming the second Task has the same name), but a new ID. So if I understand you correctly, we have auto-versioning :)

DilapidatedDucks58 trains-agent adds the artifactory URL as --extra-index-url , are you sure you are getting the correct torch version in the container? because the torch html is not an artifactory html, it is a list of links, I just want to make sure you are getting the correct version, because otherwise it can default to the CPU version, which we don't want 🙂 anyhow you can use the direct link in the "installed packages and just put there " https://download.pytorch.org/whl/nightly/cu101...

@<1571308003204796416:profile|HollowPeacock58> seems like an internal issue copying this object config.model
This is a complex object, and it seems that for some reason
None

As a workaround just do not connect this object. it seems you cannot pickle it / copy it (see GH issue)

Hmm, I think you should use --template-yaml

It uses only one CPU core, could I use multiprocessing somehow?

Hi EcstaticMouse10
Hmm, yes it should be multi core:
https://github.com/allegroai/clearml/blob/a9774c3842ea526d222044092172980ae505e24f/clearml/datasets/dataset.py#L1175
wdyt?

JitteryCoyote63

I agree that its name is not search-engine friendly,

LOL 😄
It was an internal joke the guys decided to call it "trains" cause you know it trains...
It was unstoppable, we should probably do a line of merch with AI 🚆 😉
Anyhow, this one definitely backfired...

AntsySeagull45 kudos on sorting it out 🙂
quick note, trains-agent will try to run the python version specified by the original Task. i.e. if you were running python3.7 it will first try to look for python 3.7 then if it is not there it will run the default python3. This allows a system with multiple python versions to run exactly the python version you had on your original machine. The fact that it was trying to run python2 is quite odd, one explanation I can think of is if the original e...

Ohh now I get it...
Wait a couple of hours, 0.16 is out today with trains-agent --stop flag 🙂

MassiveHippopotamus56
the "iteration" entry is actually the "max reported iteration over all graphs" per graph there is different max iteration. Make sense ?

EnviousStarfish54 generally speaking the hyper parameters are flat key/value pairs. you can have as many sections as you like, but inside each section, key/value pairs. If you pass a nested dict, it will be stored as path/to/key:value (as you witnessed).
If you need to store a more complicated configuration dict (nesting, lists etc), use the connect_configuration, it will convert your dict to text (in HOCON format) and store that.
In both cases you can edit the configuration and then when ru...

When you login with user/pass in the UI the same "process" happens and you get a Token to work with, this is the same as secret/key
Since in both cases you provide credentials and get back access token, it should work
(This is of course only if you are setting user/pass manually and disabling pass_hashed as you have)

It should work 🙂 as long as the versions match, if they don't the venv will install the version you need (which is great, only penalty is the install, download wise it will be cached)

I think I found something, let me dig deeper 🙂

Any chance you can test with the latest RC ? 1.8.4rc2

Hi DisgustedDove53

Now for the clearml-session tasks, a port-forward should be done each time if I need to access the Jupyter notebook UI for example.

So basically this is why the k8s glue has --ports-mode.
Essentially you setup a k8s service (doing the ingest TCP ports) then the template.yaml that is used by the k8s glue should specify said service. Then the clearml-session knows how to access the actual pod, by a the parameters the k8s glue sets on the Task.
Make sense ?

GiganticTurtle0 is it just --stop that throws this error ?
btw: if you add --queue default to the command line I assume it will work, the thing is , without --queue it will look for any queue with the "default" tag on it, since there are none, we get the error.
regardless that should not happen with --stop I will make sure we fix it

Just so we do not forget, can you please open an issue on clearml-agent github ?

neat! please update on your progress, maybe we should add an upgrade section once you have the details worked out

I think you are correct, it seems like it is missing requirements to boto/azure/google (I will make sure this is added). In the meantime, you can stop the "triton serving engine" Task, reset it, add boto3 to the installed packages and relaunch.
That said your main issue might be packaging the python model. Basically you need to create a model from the entire folder (with whatever there is inside the folder), then Triton should be able to run it (if the config.pbtxt is correct).
` m = OutputMo...

Could it be that this is the callback that causes it?
None

To be honest, I'm not sure I have a good explanation on why ... (unless on some scenarios an exception was thrown and caught silently and caused it)

Hi AstonishingSwan80 , what do you mean by "ec2 API"?

I added the link just in case anyway 

Smart move :)

DilapidatedDucks58 , Of course there is 🙂 actually with the latest pip 20.1 and the next RC it will be automatically detected and put into "installed package"

You can treat the "installed packages" just like you would any other "requirements.txt", just add:
git+ https://github.com/ ... and you are good to go

Hi WorriedParrot51
Let me shed some light on this complicated mechanism, because this is not very straight forward.
Basically the agent signals the trains package it should ignore the code calls, and use a specific Task in the backend (i.e. if in manual mode, the trains package logs the data into the trains-server, in agent mode (remote mode), it does the opposite and takes the data from the trains-server "into" the code)

Specifically, just like in manual mode, calling argparse.parse is be...

or shall I call the Task.init even from the agent

WorriedParrot51 I think something is lost here.
Task.init() is always called, even when the agent is executing the code. The difference is in what happens inside the Task.init() call. When the codebase itself is executed by the trains-agent, it signals through OS environment to the task.init() that instead of a new created task, it should use the already created one. from this point all data flows from the trains-server back into the c...

I mean , the python package, not the trains-server version

CloudyHamster42 what's the trains-server version ?