Fix many HPU failures in the CI #39066
more torch.hpu patches
9a93570b
increase top_k because it results in flaky behavior when Tempreture, …
732144d1
remove temporal fix
d02275b8
fix scatter operation when input and src are the same
dd7e0c79
trigger
9aae39d7
fix and reduce
d37ffc7e
skip finding batch size as it makes the hpu go loco
8990c35b
fix fsdp (yay all are passing)
0655de41
fix checking equal nan values
f30b437f
style
b74508bc
remove models list
ed974653
order
daaec7b7
rename to cuda_extensions
2db94c02
ydshieh
approved these changes
on 2025-06-30
Update src/transformers/trainer.py
6ee666c8
SunMarc
approved these changes
on 2025-07-03
SunMarc
merged
18e0cae2
into main 204 days ago
SunMarc
deleted the fix-hpu-errors branch 204 days ago
Assignees
No one assigned
Login to write a write a comment.
Login via GitHub