diff --git a/examples/train/livecodebench/lcb_dataset.py b/examples/train/livecodebench/lcb_dataset.py
index 396d5a8ec8..4016ff7e40 100644
--- a/examples/train/livecodebench/lcb_dataset.py
+++ b/examples/train/livecodebench/lcb_dataset.py
@@ -11,7 +11,7 @@
 import os
 from typing import Any, Dict, List, Optional
 
-import pandas as pd
+import datasets as hf_datasets
 
 """Utility functions for loading and processing datasets."""
 
@@ -145,17 +145,11 @@ def process_dataset(dataset_name: str, split: str, dataset_dir: str, local_dir:
     if max_rows:
         processed_data = processed_data[:max_rows]
 
-    # Save individual dataset files
-    df = pd.DataFrame(processed_data)
+    ds = hf_datasets.Dataset.from_list(processed_data)
     parquet_path = os.path.join(local_dir, f"{split}_{dataset_name}.parquet")
-    json_path = os.path.join(local_dir, f"{split}_{dataset_name}.json")
 
-    print(f"Writing {len(df)} rows to {split}_{dataset_name}.parquet")
-    df.to_parquet(parquet_path)
-
-    if split == "test":
-        print(f"Writing {len(df)} rows to {split}_{dataset_name}.json")
-        df.to_json(json_path, orient="records")
+    print(f"Writing {len(ds)} rows to {split}_{dataset_name}.parquet")
+    ds.to_parquet(parquet_path)
 
     return processed_data
 
@@ -194,7 +188,6 @@ def process_dataset(dataset_name: str, split: str, dataset_dir: str, local_dir:
     val_data = process_dataset(LIVECODEBENCH, "test", args.dataset_dir, local_dir, args.max_rows)
 
     # Save combined train dataset
-    all_train_df = pd.DataFrame(train_data)
-    all_train_df.to_parquet(os.path.join(local_dir, "deepcoder_train.parquet"))
-    all_train_df.to_json(os.path.join(local_dir, "deepcoder_train.json"), orient="records")
-    print(f"Writing {len(all_train_df)} rows to deepcoder_train.parquet and deepcoder_train.json")
+    all_train_ds = hf_datasets.Dataset.from_list(train_data)
+    all_train_ds.to_parquet(os.path.join(local_dir, "deepcoder_train.parquet"))
+    print(f"Writing {len(all_train_ds)} rows to deepcoder_train.parquet")
diff --git a/examples/train/livecodebench/run_lcb.sh b/examples/train/livecodebench/run_lcb.sh
index b2a06acf8a..f2de3d139a 100644
--- a/examples/train/livecodebench/run_lcb.sh
+++ b/examples/train/livecodebench/run_lcb.sh
@@ -5,8 +5,8 @@ set -x
 # bash examples/livecodebench/run_lcb.sh
 
 DATA_DIR="$HOME/data/lcb"
-train_data="['${DATA_DIR}/deepcoder_train.json']"
-val_data="['${DATA_DIR}/test_livecodebench.json']"
+train_data="['${DATA_DIR}/deepcoder_train.parquet']"
+val_data="['${DATA_DIR}/test_livecodebench.parquet']"
 
 # NOTE (sumanthrh): micro_train_batch_size and micro_forward_batch_size can be tuned
 uv run --isolated --frozen --extra fsdp -m skyrl.train.entrypoints.main_base \