refine Huber loss, add huber_regression_cost #3571
Merged
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
Contributor
Author
|
将book/01.fit_a_line的 |
guoshengCS
requested changes
Aug 22, 2017
paddle/gserver/layers/CostLayer.cpp
Outdated
| useGpu_ ? tmpCpuInput_[1].value->getData() : (*label.value).getData(); | ||
| std::vector<real> cost(numSamples); | ||
| for (size_t i = 0; i < numSamples; ++i) { | ||
| real a = std::abs(lbl[i] - out[i]); |
Contributor
There was a problem hiding this comment.
看上文CHECK_EQ(output.getWidth(), (*label.value).getWidth())和test_LayerGrad中config.inputDefs.push_back({INPUT_DATA, "layer_0", 10, 0})是支持多维数据的，多维情况下这里在计算时只是顺序取出了output 和label的前numSamples 个值，这样取出的数据来计算loss应该会有些问题。
| real* out = useGpu_ ? tmpCpuInput_[0].value->getData() : output.getData(); | ||
| int* lbl = useGpu_ ? tmpCpuInput_[1].ids->getData() : (*label.ids).getData(); | ||
| real* grad = useGpu_ ? tmpCpuInput_[0].grad->getData() : outputG.getData(); | ||
| for (size_t i = 0; i < numSamples; ++i) { |
Contributor
There was a problem hiding this comment.
同forwardImp ，多维情况下应该会有问题，可能因为forward和backward中取数据用了相同的方式所以test_LayerGrad可以通过。
Closed
lcy-seso
requested changes
Aug 29, 2017
paddle/gserver/layers/CostLayer.cpp
Outdated
| grad[i] += -4 * y; | ||
| else if (y * out[i] < 1) | ||
| grad[i] += -2 * (1 - y * out[i]) * y; | ||
| for (size_t j = 0; j < dim; ++j) { |
Contributor
There was a problem hiding this comment.
Huber loss 针对分类问题的这种变种只会有一维输出，不应该出现多维。分类问题不需要修改为多维。
guoshengCS
approved these changes
Aug 29, 2017
lcy-seso
approved these changes
Aug 29, 2017
|
新加的 huber_regression_cost，v1能用吗？ |
Contributor
Author
|
v1也是可以用的 |
|
需要更新 v1 的本地工具文件吗？ |
Contributor
Author
|
本地工具文件是指？需要用最新的develop分支代码 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 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.
fix #3390
Current
huber_costin Paddle is for classification, thus, rename it tohuber_classification_cost, and add an extra cost layer:huber_regression_cost.