NVIDIA / DIGITS / 924

pending completion

Build # 924

Build Type

Pull #538

travis-ci

Committed by

gheinrich

Commit Message

Re-use all weights when re-training on same dataset

Current behaviour:
When a classification model is being re-trained the last `InnerProduct` layer is renamed to allow training on another dataset (with a possibly different number of classes).
This implies that the weights of that last layer are discarded.

Problem:
If the model is being re-trained on the same dataset then initially the model will perform badly as it needs to re-learn the weights of its last `InnerProduct` layer.

Solution:
Refrain from renaming the last `InnerProduct` layer if the model is being re-trained on the same dataset

Note:
This applies to Caffe only.

closes #533

Pull Request Pull Request #538: Re-use all weights when re-training on same dataset

Run Details

7948 of 9877 relevant lines covered (80.47%)

0.8 hits per line

Source File
Press 'n' to go to next uncovered line, 'b' for previous

92.31

/digits/frameworks/torch_framework.py

Build SHA Not Found

The commit SHA "7419755fb10482fdf5e5587e41d85fe682c57233" was not found in your repository, so the file cannot be loaded. This may be because you posted from a local development environment, or your CI created an ephemeral commit.

NVIDIA / DIGITS / 924

Source File Press 'n' to go to next uncovered line, 'b' for previous

Build SHA Not Found

Source File
Press 'n' to go to next uncovered line, 'b' for previous