Don't accidentally mutate the base_model_tp_plan (#36677)
* Don't accidentally mutate the base_model_tp_plan * Co-authored by: Joao Gante <joaofranciscocardosogante@gmail.com> * Trigger tests * Marking grad accum test as slow * Add a flaky decorator * Add a flaky decorator * Use cyril's codeblock * Don't copy() when it's None * Use cyril's new codeblock * make fixup
This commit is contained in:
@@ -803,6 +803,7 @@ class TrainerIntegrationPrerunTest(TestCasePlus, TrainerIntegrationCommon):
|
||||
trainer.train()
|
||||
self.check_trained_model(trainer.model, alternate_seed=True)
|
||||
|
||||
@slow
|
||||
def test_gradient_accumulation_loss_alignment_with_model_loss(self):
|
||||
set_seed(42)
|
||||
import datasets
|
||||
|
||||
Reference in New Issue
Block a user