fix(code-quality): apply pre-commit ruff fixes

Author: ChenhanYu

examples/dataset/conversation_utils.py

@@ -186,11 +186,13 @@ def normalize_messages(example: dict[str, Any], idx: int) -> dict[str, Any]:
                 msg["tool_calls"] = m["tool_calls"]
             normalized.append(msg)
         elif role == "tool":
-            normalized.append({
-                "role": "tool",
-                "content": m.get("content") or "",
-                "tool_call_id": m.get("tool_call_id", ""),
-            })
+            normalized.append(
+                {
+                    "role": "tool",
+                    "content": m.get("content") or "",
+                    "tool_call_id": m.get("tool_call_id", ""),
+                }
+            )
         elif role == "developer":
             # Map developer-role messages to system per OpenAI schema conventions.
             normalized.append({"role": "system", "content": m.get("content") or ""})

examples/dataset/make_nemotron_ptv2_dataset.py

@@ -59,15 +59,14 @@
 import os
 from pathlib import Path
 
-from datasets import concatenate_datasets, load_dataset
-
 from conversation_utils import (
     has_tool_turns,
     load_augmentations,
     make_augment_fn,
     normalize_messages,
     strip_assistant_turns,
 )
+from datasets import concatenate_datasets, load_dataset
 
 logging.basicConfig(
     level=logging.INFO, format="%(asctime)s - %(levelname)s - %(message)s", datefmt="%H:%M:%S"

examples/dataset/make_nemotron_ptv3_dataset.py

@@ -78,15 +78,20 @@
 import argparse
 import logging
 import os
-from dataclasses import dataclass, field
+from dataclasses import dataclass
 from pathlib import Path
 from typing import Any
 
 import yaml
+from conversation_utils import (
+    has_tool_turns,
+    load_augmentations,
+    make_augment_fn,
+    normalize_messages,
+    strip_assistant_turns,
+)
 from datasets import concatenate_datasets, load_dataset
 
-from conversation_utils import has_tool_turns, load_augmentations, make_augment_fn, normalize_messages, strip_assistant_turns
-
 logging.basicConfig(
     level=logging.INFO, format="%(asctime)s - %(levelname)s - %(message)s", datefmt="%H:%M:%S"
 )
@@ -238,17 +243,14 @@ def main() -> None:
     for spec in dataset_specs:
         logger.info("Loading %s  (augment=%s)", spec.repo_id, spec.augment)
         for split in spec.splits:
-            ds = load_split(spec.repo_id, split, spec.cap_per_split, args.num_proc,
-                            args.mode)
+            ds = load_split(spec.repo_id, split, spec.cap_per_split, args.num_proc, args.mode)
             if args.mode == "generate" and not spec.augment:
                 non_augmentable_parts.append(ds)
             else:
                 augmentable_parts.append(ds)
 
     augmentable = concatenate_datasets(augmentable_parts) if augmentable_parts else None
-    non_augmentable = (
-        concatenate_datasets(non_augmentable_parts) if non_augmentable_parts else None
-    )
+    non_augmentable = concatenate_datasets(non_augmentable_parts) if non_augmentable_parts else None
     if augmentable is not None:
         logger.info("Augmentable rows: %d", len(augmentable))
     if non_augmentable is not None: