423A35C7/pytorch3d
1
0
Fork 0
mirror of https://github.com/facebookresearch/pytorch3d.git synced 2025-08-02 03:42:50 +08:00
Code Issues Packages Projects Releases Wiki Activity

nesting n_known_frames_for_test

Browse Source 
Summary: Use generator.permutation instead of choice so that different options for n_known_frames_for_test are nested.

Reviewed By: davnov134

Differential Revision: D37210906

fbshipit-source-id: fd0d34ce62260417c3f63354a3f750aae9998b0d
This commit is contained in:
Jeremy Reizenstein 2022-06-20 13:47:47 -07:00 committed by Facebook GitHub Bot
parent cba26506b6
commit 28c1afaa9d
1 changed files with 5 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
pytorch3d/implicitron/dataset/single_sequence_dataset.py
View File

@ -116,9 +116,11 @@ class SingleSceneDatasetMapProviderBase(DatasetMapProviderBase):
if set_eval_batches:
generator = np.random.default_rng(seed=0)
for batch in eval_batches:
to_add = generator.choice(
len(train_split), self.n_known_frames_for_test
)
# using permutation so that changes to n_known_frames_for_test
# result in consistent batches.
to_add = generator.permutation(len(train_split))[
: self.n_known_frames_for_test
]
batch.extend((to_add + len(split)).tolist())
split = np.concatenate([split, train_split])
frame_types.extend([DATASET_TYPE_KNOWN] * len(train_split))