Fix backward and mtl_backward bug with some tensor shapes (#227)

* Add failing parametrizations to test_value_is_correct
* Fix bug in _get_jac_matrix_chunk
* Add [YANKED] tag next to [0.4.0] header
* Add changelog entry

Author: ValerianRey

CHANGELOG.md

@@ -8,7 +8,12 @@ changes that do not affect the user.
 
 ## [Unreleased]
 
-## [0.4.0] - 2025-01-02
+### Fixed
+
+- Fixed a bug introduced in v0.4.0 that could cause `backward` and `mtl_backward` to fail with some
+  tensor shapes.
+
+## [0.4.0] - 2025-01-02 [YANKED]
 
 ### Changed
 

src/torchjd/autojac/_transform/jac.py

@@ -114,7 +114,7 @@ def _get_jac_matrix_chunk(
 
     chunk_size = jac_outputs_chunk[0].shape[0]
     if chunk_size == 1:
-        grad_outputs = [tensor.squeeze() for tensor in jac_outputs_chunk]
+        grad_outputs = [tensor.squeeze(0) for tensor in jac_outputs_chunk]
         gradient_vector = get_vjp(grad_outputs)
         return gradient_vector.unsqueeze(0)
     else:

tests/unit/autojac/test_backward.py

@@ -24,7 +24,7 @@ def test_various_aggregators(aggregator: Aggregator):
 
 
 @mark.parametrize("aggregator", [Mean(), UPGrad()])
-@mark.parametrize("shape", [(2, 3), (2, 6), (5, 8), (20, 55)])
+@mark.parametrize("shape", [(1, 3), (2, 3), (2, 6), (5, 8), (20, 55)])
 @mark.parametrize("manually_specify_inputs", [True, False])
 @mark.parametrize("chunk_size", [1, 2, None])
 def test_value_is_correct(

tests/unit/autojac/test_mtl_backward.py

@@ -27,7 +27,7 @@ def test_various_aggregators(aggregator: Aggregator):
 
 
 @mark.parametrize("aggregator", [Mean(), UPGrad()])
-@mark.parametrize("shape", [(2, 3), (2, 6), (5, 8), (20, 55)])
+@mark.parametrize("shape", [(1, 3), (2, 3), (2, 6), (5, 8), (20, 55)])
 @mark.parametrize("manually_specify_shared_params", [True, False])
 @mark.parametrize("manually_specify_tasks_params", [True, False])
 @mark.parametrize("chunk_size", [1, 2, None])