Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Pairwise Noise Contrastive Alignment #1632

Merged
merged 7 commits into from
May 14, 2024
Merged

Pairwise Noise Contrastive Alignment #1632

Changes from 1 commit
Commits
Show all changes
7 commits
Select commit Hold shift + click to select a range
8891a1d
add NCA paired preference loss
winglian May 6, 2024
457179d
chore: lint
winglian May 10, 2024
f709dc9
set more lenient tolerance for integration tests
winglian May 10, 2024
770cf33
Update tests/test_dpo_trainer.py
younesbelkada May 14, 2024
1f036d2
skip test
younesbelkada May 14, 2024
e255664
merge
younesbelkada May 14, 2024
1f7038b
fix
younesbelkada May 14, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Prev Previous commit
Next Next commit
Update tests/test_dpo_trainer.py
  • Loading branch information
@younesbelkada
younesbelkada authored May 14, 2024
commit 770cf333f6cfd3241ab63d3b51a00abef93eac43
2 changes: 1 addition & 1 deletion tests/test_dpo_trainer.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -142,7 +142,7 @@ def test_dpo_trainer(self, name, loss_type, pre_compute):
new_param = trainer.model.get_parameter(n)
# check the params have changed - ignore 0 biases
if param.sum() != 0:
assert not torch.allclose(param, new_param, rtol=1e-12, atol=1e-12)
assert not torch.allclose(param, new_param, rtol=1e-16, atol=1e-16)

def test_dpo_trainer_without_providing_ref_model(self):
with tempfile.TemporaryDirectory() as tmp_dir:
Expand Down
Loading