QC Plugin: Remove handling of triple quoted transcripts

skit-ai · Jul 2, 2024 · 06a8db8 · 06a8db8
1 parent e5b8e34
commit 06a8db8
Show file tree

Hide file tree

Showing 2 changed files with 0 additions and 13 deletions.
diff --git a/dialogy/plugins/text/qc_plugin/__init__.py b/dialogy/plugins/text/qc_plugin/__init__.py
@@ -53,9 +53,6 @@ def identify_conflicting_labels(training_data: pd.DataFrame) -> pd.DataFrame:
 
         logger.debug(f"Finding data points with conflicting labels...")
 
-        training_data["alternatives"] = training_data["alternatives"].apply(
-            lambda x: x.replace("""\"\"""", """\"""") if isinstance(x, str) else x
-        )
         training_data["frozen_set_hash"] = training_data["alternatives"].apply(
             lambda x: hashlib.md5(
                 pickle.dumps(

diff --git a/tests/plugin/text/test_qc_plugin/test_qc_plugin.py b/tests/plugin/text/test_qc_plugin/test_qc_plugin.py
@@ -40,16 +40,6 @@
             True,
             2,
         ),
-        (
-            [
-                '[[{"transcript": "hello"}]]',
-                '[[{"transcript": "hello"}]]',
-                """[[{\""confidence\"": 0.801317, \""transcript"" :\""hello\""}]]""",
-            ],
-            ["x1", "x2", "x3"],
-            True,
-            3,
-        ),
     ],
 )
 async def test_drop_conflicting_labels(alternatives, tags, drop, discard_size, tmp_path) -> None: