I get an error when i try to press "Training"

[AntoineMath]

I followed the video tutorial and when i want to train i get this error :
RuntimeError: Task container finished with non-zero status: {'Error': None, 'StatusCode': 134}

[AntoineMath]

Fixed : my batch size was to big, i have a poor GTX970M and i set batch size to "1" ^^

[sushantgundla]

Even after setting batch size to 1 I'm getting the same error.
I have Quadro K260 graphics card.

[bakhtiar-afbl]

i also face same problem--
i am using GeForce RTX 2070 Graphics Card
Error: Task container finished with non-zero status: {'Error': None, 'StatusCode': 134}
please give any solution

[danieltwx]

@AntoineMath @bakhtiar-afbl @Sharpyyy
Hi guys, im facing the same issue when using supervisely. Would be glad to hear what solutions did u all made to solve the issue!

[sushantgundla]

Did not get any solutions yet!

[ikomhoog]

Same problem, did anyone find a solution yet?

[aparico]

The same thing happened to me when I set "gpu_devices" to [ 1 ] in the config file.

But apparently, it should be
"gpu_devices": [ 0 ]

because [ 0 ] is an index to indicate one GPU device. Then, it worked for me.

[fourcolors]

I'm getting the same issue with a GTX970

[97Fekim]

Fixed : my batch size was to big, i have a poor GTX970M and i set batch size to "1" ^^

Did you fix this error by downsizing batch size???!!

[AbdullahJirjees]

@bakhtiar-afbl I am using same your gpu and I am getting the same error number, my batch size is Training : 3 and val: 3 any idea how to solve it?

[AbdullahJirjees]

@AntoineMath @Sharpyyy had you manged to solve the error?

[SushantGundla-Draup]

@AbdullahAmer86 Not yet