Update conditional logic and -> or in SAM postprocessing#23295
Conversation
|
The documentation is not available anymore as the PR was closed or merged. |
|
@younesbelkada exactly but with the current implementation it could also be a list of arrays and the exception would not be thrown |
younesbelkada
left a comment
younesbelkada
left a comment
There was a problem hiding this comment.
Ah yes, you are right, that sounds correct to me! Thanks for the heads up
amyeroberts
changed the title
|
FYI, this PR triggered 3 test failures on the CI. (The third one is GPU OOM, which is likely caused by the other 2 failures). tests/models/sam/test_modeling_sam.py::SamModelIntegrationTest::test_inference_mask_generation_one_point_one_bb
(line 235) ValueError: Input boxes must be a list of list of list of floating integers.
tests/models/sam/test_modeling_sam.py::SamModelIntegrationTest::test_inference_mask_generation_one_point_one_bb_zero
(line 235) ValueError: Input boxes must be a list of list of list of floating integers.
tests/models/sam/test_modeling_sam.py::SamModelIntegrationTest::test_inference_mask_generation_two_points_batched
(line 808) torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 2.00 GiB (GPU 0; 14.76 GiB total capacity; 11.65 GiB already allocated; 792.75 MiB free; 12.74 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF |
or or and
or or and
|
ValueError: Input boxes must be a list of list of list of floating integers. I AM Still getting this error, when even run the example notebooks, what do I do? I just have a list of bounding box coordinated(flatenned) |
|
Hi @karthikdatta98 |
6 participants
should be ors here