fastxml icon indicating copy to clipboard operation
fastxml copied to clipboard

Building module fastxml.inferencer failed: ["distutils.errors.CompileError: command 'gcc' failed with exit status 1\n"]

Open YuMiaoTHU opened this issue 6 years ago • 5 comments

when I import the inferencer.pyx, it is wrong? How can I solve it? Thanks!

YuMiaoTHU avatar Jul 10 '19 07:07 YuMiaoTHU

I need a bit more to go on. Can I get your system info, gcc version, and full stack trace?

Refefer avatar Jul 13 '19 03:07 Refefer

tanks for your reply! I have fixed the error, by set_up.py build_ext --inplace but there are another problem!!! when I use my own data to train the model, 500,000 data, when train after almost 6 hours, the code will be killed! I use cpu to train the model. the command is fxml.py my_json.model train.json --verbose train --iters 5 --trees 20 --label-weight propensity --alpha 1e-4 --leaf-classifiers I don't konw why........... thanks

YuMiaoTHU avatar Jul 13 '19 06:07 YuMiaoTHU

sorry, I think I should add more details. That is I use all cpu in a linux sever, when I use 1,000 data, There is no problem in the training process, but when I use 500,000 data to train, The process took about 6 hours and got stuck there. It didn't show killed, but the process was gone. like this, Splitting 1564 Training classifier Training classifier Has been showing for more than 24 hours, I suspect it is a memory leak, but I don't know how to fix it.

image the process is showed below image thanks!

YuMiaoTHU avatar Jul 15 '19 10:07 YuMiaoTHU

My hunch is you ran out of memory on your machine. How big is the training dataset and how much memory is in the machine? How many features in the dataset?

On Mon, Jul 15, 2019, 6:55 AM YuMiaoTHU [email protected] wrote:

sorry, I think I should add more details. That is I use all cpu in a linux sever, when I use 1,000 data, There is no problem in the training process, but when I use 500,000 data to train, The process took about 6 hours and got stuck there. It didn't show killed, but the process was gone. like this,

Splitting 1564 Training classifier Training classifier Has been showing for more than 24 hours, I suspect it is a memory leak, but I don't know how to fix it.

[image: image] https://user-images.githubusercontent.com/42285896/61211177-76360a80-a731-11e9-9e07-00d2ec861056.png the process is showed below [image: image] https://user-images.githubusercontent.com/42285896/61211376-0411f580-a732-11e9-8cbb-528f48528aa1.png thanks!

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/Refefer/fastxml/issues/23?email_source=notifications&email_token=AALUUFIB72J5OBGOGM2MNIDP7RJQTA5CNFSM4H7MPCMKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODZ5K5SI#issuecomment-511356617, or mute the thread https://github.com/notifications/unsubscribe-auth/AALUUFJZLBPEVMLFGM6JYXDP7RJQTANCNFSM4H7MPCMA .

Refefer avatar Jul 16 '19 22:07 Refefer