Merge pull request #6269 from rapidsai/branch-25.02

Forward-merge branch-25.02 into branch-25.04

cpp/src/fil/fil.cu

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2024, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2025, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -391,6 +391,14 @@ struct dense_forest<dense_node<real_t>> : forest<real_t> {
   {
     this->init_common(h, cat_sets, vector_leaf, params);
     if (this->algo_ == algo_t::NAIVE) this->algo_ = algo_t::BATCH_TREE_REORG;
+    /* The following line is added to disable BATCH_TREE_REORG mode. Currently
+     * there is a bug in the implementation of this mode which causes a batch
+     * of trees to be marked as complete before every tree has reached a leaf
+     * node. At best, this results in a crash due to a bad memory access, but
+     * it can also silently impact the final output of the model. No issue has
+     * been logged for this workaround because we will be migrating to the new
+     * FIL codebase which is not impacted by this bug. */
+    if (this->algo_ == algo_t::BATCH_TREE_REORG) { this->algo_ = algo_t::TREE_REORG; }
 
     int num_nodes = forest_num_nodes(this->num_trees_, this->depth_);
     nodes_.resize(num_nodes, h.get_stream());

python/cuml/cuml/benchmark/bench_helper_funcs.py

@@ -1,5 +1,5 @@
 #
-# Copyright (c) 2019-2024, NVIDIA CORPORATION.
+# Copyright (c) 2019-2025, NVIDIA CORPORATION.
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
 # you may not use this file except in compliance with the License.
@@ -142,7 +142,7 @@ def _build_fil_classifier(m, data, args, tmpdir):
     num_rounds = args["num_rounds"]
     n_feature = data[0].shape[1]
     train_size = data[0].shape[0]
-    model_name = f"xgb_{max_depth}_{num_rounds}_{n_feature}_{train_size}.model"
+    model_name = f"xgb_{max_depth}_{num_rounds}_{n_feature}_{train_size}.ubj"
     model_path = os.path.join(tmpdir, model_name)
     bst = xgb.train(params, dtrain, num_rounds)
     bst.save_model(model_path)