Closed The-Mr-L closed 6 months ago
btw the model setup is like so
let mut model = GradientBooster::default()
.set_iterations(151)
.set_objective_type(forust_ml::objective::ObjectiveType::LogLoss)
.set_learning_rate(0.01)
.set_max_depth(5)
.set_gamma(0.)
.set_initialize_base_score(false)
.set_base_score(0.5)
.set_min_leaf_weight(1.)
.set_l2(1.);
model.fit_unweighted(&matrix, &y, None)?;
model.save_booster("./model.json")?;
Thanks for reporting this! I see the problem, will close this when the fix is released.
Please try again with version 0.4.5.
it works now :) thanks !
Hi :) just a quick heads up, I have not had the time yet to troubleshoot it , but I can without any problems train a model from python and load it from both python and rust but the trained model from rust can only be loaded from rust. well the error is just a panic like this "PanicException: called
Option::unwrap()
on aNone
value" . I will report back if I find the problem