Repository navigation
FastTree: FastTreeRanking per-iteration loss metrics are empty #242
Description
System information
- OS version/distro: Various Windows 10 machines, Windows Server 2012 or 2016.
- .NET Version (eg., dotnet --info): 4.7, 4.6.x
Although, not a system-related bug.
Issue
- What did you do? Train on a dataset using
FastTreeRankingwith thetestFrequencyparameter set to something like 1. Optionally, add a validation set. FastTree will now report training metrics on the train (and validation) sets every N iterations. Do the same forFastTreeRegressionandFastTreeClassification. - What happened? For
RegressionandClassification, the expected output is printed. ForRanking, The header for the per-iteration test and validation metrics prints, but there are no metrics. - What did you expect? I expected metrics to be printed for the
Rankingcase.
Source code / logs
Sample output:
trainNDCG:
validNDCG:
trainNDCG:
validNDCG:
trainNDCG:
validNDCG:
trainNDCG:
validNDCG:
Activity
rogancarr commentedon May 25, 2018
I have done a preliminary investigation, and for at least the
ranking-sampledataset (which we need to add to this repository), the root cause is that thelengthof theDataset.MaxDCGproperty is0. Since the length is 0, no scores are computed. It looks like the MaxDCG vector is initialized at double[0][] at the beginning of training and never updated. My working hypothesis is that this is only set during scoring.Next steps:
ML.NET-242: FastTreeRanking per-iteration loss metrics are empty
ML.NET-242: FastTreeRanking per-iteration loss metrics are empty
ML.NET-242: FastTreeRanking per-iteration loss metrics are empty (#289)
Merge master into release/preview branch for v0.3 (#426)
ML.NET-242: FastTreeRanking per-iteration loss metrics are empty (dot…
Samples / API rev from 2/27 bug bash feedback (dotnet#242)
code migration from machinelearning-automl to machinelearning repo fe…
Samples / API rev from 2/27 bug bash feedback (dotnet#242)
Added onnx export functionality for LpNormNormalizingTransformer (#4161)
Added Onnx export functionality to PCATransformer (#4188)