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

Jump to bottom

Update LearningRate for test fit a line BF16 #34653

Merged
merged 4 commits into from
Aug 24, 2021
Merged

Update LearningRate for test fit a line BF16 #34653

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
ca83726
Small corrections.
arogowie-intel Aug 5, 2021
89e4d3d
Fix lr for bf16.
arogowie-intel Aug 5, 2021
a843a50
Revert some changes.
arogowie-intel Aug 23, 2021
fc61de5
Merge remote-tracking branch 'upstream/develop' into aosewski/fit_a_line
arogowie-intel Aug 23, 2021
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
2 changes: 1 addition & 1 deletion python/paddle/fluid/contrib/mixed_precision/bf16/amp_utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -235,7 +235,7 @@ def bf16_guard():
def are_post_ops_bf16(post_ops, keep_fp32_ops):
for post_op in post_ops:
for op in post_op:
if op.type in keep_fp32_ops:
if op in keep_fp32_ops:
return False
return True

Expand Down
5 changes: 3 additions & 2 deletions python/paddle/fluid/tests/book/test_fit_a_line.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -48,7 +48,8 @@ def train(use_cuda, save_dirname, is_local, use_bf16, pure_bf16):
cost = fluid.layers.square_error_cost(input=y_predict, label=y)
avg_cost = fluid.layers.mean(cost)

sgd_optimizer = fluid.optimizer.SGD(learning_rate=0.001)
lr = 5e-3 if use_bf16 else 1e-3
sgd_optimizer = fluid.optimizer.SGD(learning_rate=lr)

if use_bf16:
sgd_optimizer = amp.bf16.decorate_bf16(
Expand Down Expand Up @@ -83,7 +84,7 @@ def train_loop(main_program):
avg_loss_value, = exe.run(main_program,
feed=feeder.feed(data),
fetch_list=[avg_cost])
if avg_loss_value[0] < 10.0 or pure_bf16:
if avg_loss_value[0] < 10.0:
if save_dirname is not None:
paddle.static.save_inference_model(save_dirname, [x],
[y_predict], exe)
Expand Down