[Accuracy diff No.91] Fix accuracy diff for paddle.incubate.nn.functional.fused_layer_norm API #343
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
Thanks for your contribution! |
ooooo-create
force-pushed
the
accuracy_fused_layer_norm
branch
from
f61cc7f to
585bf5d
Compare
ooooo-create
force-pushed
the
accuracy_fused_layer_norm
branch
from
585bf5d to
66af659
Compare
Collaborator
|
@ooooo-create 麻烦o师傅解决一下冲突～ |
ooooo-create
force-pushed
the
accuracy_fused_layer_norm
branch
from
66af659 to
55a9adb
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
删除非法配置
paddle.incubate.nn.functional.fused_layer_norm(Tensor([2, 64],"float16"), norm_weight=Tensor([64],"float32"), norm_bias=Tensor([64],"float32"), epsilon=1e-05, begin_norm_axis=1, bias=Tensor([64],"float16"), residual=Tensor([2, 1, 64],"float16"), )paddle.incubate.nn.functional.fused_layer_norm(Tensor([16, 256],"float16"), Tensor([256],"float32"), Tensor([256],"float32"), 1e-05, begin_norm_axis=1, bias=Tensor([256],"float16"), residual=Tensor([16, 256],"float16"), residual_alpha=0.69204696, quant_scale=0.15, quant_round_type=1, quant_max_bound=127, quant_min_bound=-127, )AMP 出现精度误差，参考 #389 添加atol = 1（因为在量化时涉及 round() 操作会四舍五入）手动改成 float32 是没有精度问题其余测试均通过