Mercurial > pylearn
comparison doc/v2_planning/dataset.txt @ 1110:4797a4cb73e1
added comment to dataset.
| author | Frederic Bastien <nouiz@nouiz.org> |
|---|---|
| date | Tue, 14 Sep 2010 09:08:11 -0400 |
| parents | 29b48deb6a84 |
| children | 18a092001752 |
comparison
equal
deleted
inserted
replaced
| 1109:29b48deb6a84 | 1110:4797a4cb73e1 |
|---|---|
| 355 For the learner to do this ( since it gets the same interface from any | 355 For the learner to do this ( since it gets the same interface from any |
| 356 dataset object) would be like and if <this case> then, while for the | 356 dataset object) would be like and if <this case> then, while for the |
| 357 dataset is just a different class. But I'm happy to have all this GPU stuff | 357 dataset is just a different class. But I'm happy to have all this GPU stuff |
| 358 send to the learner as well if everybody else believe that is better. | 358 send to the learner as well if everybody else believe that is better. |
| 359 | 359 |
| 360 FB comment: I don't understand why you would need to recompile the theano function. | |
| 361 Their is 2 cases, the data is in a shared variable. You can directly change the data | |
| 362 in the shared variable without recompiling the theano fct. The second case is when | |
| 363 the dataset is in an ordinary theano variable. In that case, the first step in the | |
| 364 theano fct will be to transfer the dataset to the gpu before computation. If the data | |
| 365 change at each call, that will be as efficient as changing the data manually every time | |
| 366 in the shared variable. |