[MRG+1] DOC: updating GridSearchCV's n_jobs parameter #8106
Conversation
|
|
||
| n_jobs: int, default: 1 : | ||
| The maximum number of concurrently running jobs, such as the number of | ||
| Python worker processes when backend=”multiprocessing” or the size of |
There was a problem hiding this comment.
The user cannot really choose the backend here so this is too much information, I think.
Perhaps copy an n_jobs description from elsewhere in scikit-learn, rather than from joblib.parallel.
Also, please replicate your changes to RandomizedSearchCV.
ec0111c to
54bf333
Compare
|
@jnothman thanks for the feedback, I made the necessary changes. Let me know if anything else needs to be fixed. |
| n_jobs : int, default=1 | ||
| Number of jobs to run in parallel. | ||
| n_jobs: int, default: 1 : | ||
| The maximum number of concurrently running jobs. |
There was a problem hiding this comment.
Could be more explicit, "estimators fit in parallel"
54bf333 to
5131d3c
Compare
|
Wait: doesn't this still need to be applied to RandomizedSearchCV? |
|
@jnothman it was added to |
|
Sorry for my slowness. LGTM |
|
|
||
| - For n_jobs below -1, (n_cpus + 1 + n_jobs) are used. | ||
|
|
||
| - For n_jobs = -2, all CPUs but one are used. |
There was a problem hiding this comment.
I think it's better to merge the last two points. Maybe something like :
For ``n_jobs`` below -1, ``(n_cpus + n_jobs + 1)`` are used. For example, with ``n_jobs = -2`` all CPUs but one are used.
What do you think ?
eca4339 to
c2e48e0
Compare
c2e48e0 to
88c63d5
Compare
|
thanks for the review @tguillemot , I made the change but not sure why one of the appveyor tests is failing now.... |
|
Clearly not your fault. Merging, thanks! |
|
Thx @accraze. |
Reference Issue
Fixes #5797
What does this implement/fix? Explain your changes.
Clarify n_jobs for gridsearch in docs
Any other comments?