Error at training

[alvaro.marlo (Álvaro Martínez López)]

Hi,

First of all I don't know if this is a bug related to Prodigy, but I would like to know if you can give me a hand.

When I run the training, it stops and returns this message:

{route-to-my-project-env}/bin/prodigy: line 2: 22878 Killed: 9   python -m prodigy "$@"

Any idea what it could be due to?

Thanks,
Álvaro

[ryanwesslen]

hi @alvaro.marlo!

@alvaro.marlo

When I run the training, it stops and returns this message:

{route-to-my-project-env}/bin/prodigy: line 2: 22878 Killed: 9   python -m prodigy "$@"

Any idea what it could be due to?

"Killed" typically is indicative of memory issues. Especially since you're training, can you try to train on a smaller dataset?

Can't install Prodigy: "Killed" during pip installation

I’m trying to upgrade from Prodigy 1.6.1 to 1.7.1 in a virtual machine. I keep getting a ‘killed’ message. To try to increase disk and memory space, I uninstalled the old version, but I still can’t get the new one installed! This is the error message: (env) xxxxx@xxxxx:~$ pip install prodigy-1.7.1-cp35.cp36.cp37-cp35m.cp36m.cp37m-linux_x86_64.whl Processing ./prodigy-1.7.1-cp35.cp36.cp37-cp35m.cp36m.cp37m-linux_x86_64.whl Requirement already satisfied: ujson==1.35 in ./env/lib/python3.5/site-…

Depending on your OS, there are a few tools for Windows, Mac or Linux to track memory.

Be sure to also avoid any very long documents. That can sometimes cause memory issues when training.

Last, sometimes using the data-to-spacy recipe and its outputs to train with spacy train may help:

Prodigy NER train recipe getting killed for no apparent reason

It's hard to say for sure, but given that one dataset less works fine I'm indeed guessing it's a memory issue. Are you able to export the datasets to the .spacy format via the data-to-spacy recipe? If so, we might be able to pick it up from spaCy.