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

[AutoParallel] fix fp16 for subblock #47189

Merged
merged 3 commits into from
Oct 20, 2022
Merged

[AutoParallel] fix fp16 for subblock #47189

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
97ddbc8
[AutoParallel] fix fp16 for subblock
Oct 19, 2022
9cdcc89
fix engine
Oct 19, 2022
9199af2
fix comment
Oct 19, 2022
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
6 changes: 3 additions & 3 deletions python/paddle/distributed/auto_parallel/engine.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -243,8 +243,8 @@ def _infer_item_spec(item, name, batch_size, specs):
else:
specs.append(spec.batch(batch_size))
elif isinstance(item, (Variable, core.VarBase, core.eager.Tensor)):
_adjust_item_spec(num_shards, spec)
spec = InputSpec.from_tensor(item, name)
_adjust_item_spec(num_shards, spec)
if batch_size is None:
specs.append(spec)
else:
Expand Down Expand Up @@ -1530,10 +1530,10 @@ def load(self, path, strict=True, load_optimizer=True):
strict (bool, optional): Whether to skip the loading of mismatch
parameter or raise an error when mismatch happens (not found
the parameter in file storing model states of or receives a
mismatch shape). Default: False.
mismatch shape). Default: True.
load_optimizer (bool, optional): If True, the stored optimizer
states is restored. Otherwise, the optimizer states is initialized
from scratch. Default: False.
from scratch. Default: True.

Returns:
None
Expand Down
3 changes: 2 additions & 1 deletion python/paddle/distributed/passes/auto_parallel_fp16.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -181,7 +181,8 @@ def set_var_to_fp16(self, var_name, block):
try:
var = block.var(var_name)
except ValueError as e:
var = self.program.global_block().var(var_name)
var = block._var_recursive(var_name)
# var = self.program.global_block().var(var_name)

# NOTE(JZ-LIANG) "array_" is a hack to adopt for ernie3.0 inference, since there is
# a trick which make the LOD_TENSOR_ARRAY to the float32 in while block to reset the LOD_TENSOR_ARRAY
Expand Down