From bf5162805a7a18e2629c03cdfc8a994597e73b3a Mon Sep 17 00:00:00 2001 From: Maxime Gasse Date: Wed, 30 Oct 2024 13:30:54 -0400 Subject: [PATCH] rename benchmark after subset_from_split() --- .../experiments/src/browsergym/experiments/benchmark/base.py | 1 + 1 file changed, 1 insertion(+) diff --git a/browsergym/experiments/src/browsergym/experiments/benchmark/base.py b/browsergym/experiments/src/browsergym/experiments/benchmark/base.py index d4261be8..8ee90407 100644 --- a/browsergym/experiments/src/browsergym/experiments/benchmark/base.py +++ b/browsergym/experiments/src/browsergym/experiments/benchmark/base.py @@ -142,6 +142,7 @@ def subset_from_split(self, split: Literal["train", "valid", "test"]): # recover the target split sub_benchmark = self.subset_from_regexp(split_column, regexp=f"^{split}$") + sub_benchmark.name = f"{self.name}_{split}" # check that the split exists (non-empty task list) if not sub_benchmark.env_args_list: