MagnificentPig49 I was not aware of jsonargparse from what I understand it's a nicer way to parse json configuration files, with argparser alike interface. Did I get that correctly?
Regrading the missing argparser, you are correct, the auto-magic is not working since jsonargparse is calling an internal ArgParser function and not the external one (hence we miss it).
The quickest fix is adding the following line before you call parse_args() :
task.connect(parent_parser)

AgitatedDove14 thanks but I’m not using argparse , I’m using jsonargparse
you can see the simplified snippet here
from jsonargparse import ArgumentParser from trains import Task model_snapshots_path = 'mnt/trains' task = Task.init(project_name='examples', task_name='test argparser', output_uri=model_snapshots_path) logger = task.get_logger() def main(args): print('Got args: %s' % args) if __name__ == '__main__': parent_parser = ArgumentParser(add_help=False) parent_parser.add_argument('--num_workers', help='integer value', type=int, default=1) parent_parser.add_argument('--gpus', help='integer value', type=int, default=0) parent_parser.add_argument('--max_epochs', help='integer value', type=int, default=1) parent_parser.add_argument('--visdom', action='store_true') main(parent_parser.parse_args())

Thanks AgitatedDove14 yes task.connect worked. about jsonargparse not sure how popular it is but in our case regular argparse didn’t work and we find out jsonargparse is working perfect. I guess will be beneficial to support that as well.

thanks SuccessfulKoala55 , sorry that I can’t share the whole code but here is my main part
` import faulthandler
import logging

import torch
from argparse import ArgumentParser
from pytorch_lightning.callbacks import ModelCheckpoint

...

import os

from trains import Task
model_snapshots_path = 'mnt/trains'
if not os.path.exists(model_snapshots_path):
os.makedirs(model_snapshots_path)

task = Task.init(project_name='xl_trainer',
task_name='extending automagical Trains xl trianer',
output_uri=model_snapshots_path)

logger = task.get_logger()

def main(args):

 ` ` ('Got args: %s', args)

model = XLModelFactory().create_or_load(args)

...

if name == 'main':
parent_parser = ArgumentParser(add_help=False)
main(ResnetCIFAR10Model.add_model_specific_args(parent_parser).parse_args(nested=True)) `

MagnificentPig49 I'll look into it, if it's very simple, it should not be a problem.

MagnificentPig49 is jsonargparse popular, should we fix the auto-magic ?

I simplified the code, just so I could test it, this one seems to work, feel free to add the missing argparser parts :)
` from argparse import ArgumentParser
from trains import Task
model_snapshots_path = 'mnt/trains'
task = Task.init(project_name='examples', task_name='test argparser', output_uri=model_snapshots_path)
logger = task.get_logger()

def main(args):
print('Got args: %s' % args)

if name == 'main':
parent_parser = ArgumentParser(add_help=False)
parent_parser.add_argument('--argparser_int_value', help='integer value', type=int, default=1)
main(parent_parser.parse_args()) `

Hi MagnificentPig49 ,
Can you share a little bit more of your initial code? Where exactly do you call http://Task.in

Hi MagnificentPig49 I suspect it's the ResnetCIFAR10Model wrapper, is it your module ?