Testing with different number of rounds about xgboost HOT 4 CLOSED

Please refer this:
https://github.com/tqchen/xgboost/blob/master/demo/guide-python/predict_first_ntree.py

Sent from Bing's Nexus 5
On Sep 14, 2014 8:32 AM, "davutpolat" [email protected] wrote:

Hi,

i trained a model with 2500 rounds, it took hours to train it, i also
wonder result of model which is trained with 2000 or less round, but it
will take hours again, i want to use this model (with 2500 rounds) but want
test step to consider its first 2000 rounds. does xgboost support to test a
model with different number of rounds?

Regards,
Davut

—
Reply to this email directly or view it on GitHub
https://github.com/tqchen/xgboost/issues/77.

from xgboost.

Thanks for quick answer, i pulled xgboost about 1 month ago, and still use same version, does it support this feature?

from xgboost.

Please pull the newest.

Sent from Bing's Nexus 5
On Sep 14, 2014 8:45 AM, "davutpolat" [email protected] wrote:

Thanks for quick answer, i pulled xgboost about 1 month ago, and still use
same version, does it support this feature?

—
Reply to this email directly or view it on GitHub
https://github.com/tqchen/xgboost/issues/77#issuecomment-55527772.

from xgboost.

This is a recent feature in .32 so it is unlikely your versio support it

On Sunday, September 14, 2014, davutpolat [email protected] wrote:

Bing,

i will but i am still training a model right now, i have been waiting for
6 hours to see its result, i cant stop it, if my version does not support
it will stop right now, when was this feature commited?

Regards,

—
Reply to this email directly or view it on GitHub
https://github.com/tqchen/xgboost/issues/77#issuecomment-55527936.

from xgboost.