Style changes

h2oai · Sep 20, 2024 · 43742d0 · 43742d0
1 parent 8805ff6
commit 43742d0
Show file tree

Hide file tree

Showing 2 changed files with 12 additions and 9 deletions.
diff --git a/llm_studio/app_utils/sections/dataset.py b/llm_studio/app_utils/sections/dataset.py
@@ -706,12 +706,11 @@ async def dataset_import(
             sanity_check(cfg)
 
         except AssertionError as exception:
-            logger.error(
-                f"Error while validating data: {exception}", exc_info=True
-            )
+            logger.error(f"Error while validating data: {exception}", exc_info=True)
             text = (
                 "# Error while validating data\n"
-                "Please go back and verify whether the problem type and other settings were set properly.\n"
+                "Please go back and verify whether the problem type and other "
+                "settings were set properly.\n"
                 "\n"
                 "**Details of the Validation Error**:\n"
                 f"```\n{exception}\n```"

diff --git a/llm_studio/src/datasets/text_causal_language_modeling_ds.py b/llm_studio/src/datasets/text_causal_language_modeling_ds.py
@@ -336,18 +336,22 @@ def sanity_check(cls, df: pd.DataFrame, cfg: Any, mode: str = "train"):
             and cfg.dataset.parent_id_column in df.columns
             and "id" in df.columns
         ):
-            assert (
-                df[cfg.dataset.parent_id_column] != df["id"]
-            ).all(), f"Parent id column:{cfg.dataset.parent_id_column} is the same as id column for some rows"
+            assert (df[cfg.dataset.parent_id_column] != df["id"]).all(), (
+                f"Parent id column:{cfg.dataset.parent_id_column}"
+                "is the same as id column for some rows"
+            )
             assert (df[cfg.dataset.parent_id_column].fillna("") == "").sum() > 0, (
                 "Did not find any conversation chain. "
                 "Please ensure that some parent ids are empty."
                 "\n"
-                "Conversations are chained using parent id, start conversation record should not have parent id populated"
+                "Conversations are chained using parent id, "
+                "start conversation record should "
+                "not have parent id populated"
                 "\n"
                 f"Parent id column checked:{cfg.dataset.parent_id_column}"
                 "\n"
-                f"Number of records with empty <{cfg.dataset.parent_id_column}> column:{(df[cfg.dataset.parent_id_column].fillna('') == '').sum()}"
+                f"Number of records with empty <{cfg.dataset.parent_id_column}>"
+                "column:{(df[cfg.dataset.parent_id_column].fillna('') == '').sum()}"
             )
 
         assert cfg.dataset.answer_column in df.columns, (