ProphetStor models are replaced with in-house developed models.
Preprocessors are also stored in addition to the prediction models.
Objects are now stored using joblib instead of pickle, as recommended by
scikit-learn docs.
"manufacturer-specific" models are used instead of "best-feature-match"
models. i.e., instead of models being trained (presumably) just based on what
features are available, models have been trained for each manufacturer.
This is because of variation in meaning and availibility of SMART
attributes across manufacturers.
Updated config.json, requirements.txt, and DiskFailurePredictor for these changes.
new file mode 100644 (file)
index 0000000..9e1c51f
Binary files /dev/null and b/src/pybind/mgr/diskprediction_local/models/hgst_predictor.joblib differ
new file mode 100644 (file)
index 0000000..2d94963
Binary files /dev/null and b/src/pybind/mgr/diskprediction_local/models/hgst_preprocessor.joblib differ
new file mode 100644 (file)
index 0000000..574223e
Binary files /dev/null and b/src/pybind/mgr/diskprediction_local/models/seagate_predictor.joblib differ
new file mode 100644 (file)
index 0000000..34f96ab
Binary files /dev/null and b/src/pybind/mgr/diskprediction_local/models/seagate_preprocessor.joblib differ