diff --git a/giant/analysis.py b/giant/analysis.py
index 0a5bd01..b32b992 100644
--- a/giant/analysis.py
+++ b/giant/analysis.py
@@ -105,11 +105,10 @@ from giant.constants import (
     PREDICT_SCHEMA_VERSION,
     PREDICT_SCHEMA_VERSION_KEY,
 )
-from giant.data.dataset import train_val_split
-from giant.data.loader import find_parquet_files, load_steps
+from giant.data.dataset import make_event_split, StreamingStepsDataset
+from giant.data.loader import find_parquet_files, load_event_ids
 from giant.data.transforms import (
     Normalizer,
-    build_features,
     energy_simplex_decode,
     inv_log_transform,
     reconstruct_post_pos,
@@ -275,20 +274,20 @@ def make_val_loader(
     very large datasets, build a StreamingStepsDataset directly (see giant.pipeline).
     """
     files = find_parquet_files(data)
-    chunks = [load_steps(f) for f in files]
-    full = {k: np.concatenate([c[k] for c in chunks]) for k in chunks[0]}
+    all_event_ids = np.concatenate([load_event_ids(f) for f in files])
+    _, val_events = make_event_split(all_event_ids, val_fraction=val_fraction, seed=seed)
 
-    cond_cont, cond_cat, target, _, _ = build_features(
-        full,
-        bundle.pdg_map,
-        bundle.mat_map,
+    val_ds = StreamingStepsDataset(
+        files=files,
+        split_events=val_events,
+        pdg_map=bundle.pdg_map,
+        mat_map=bundle.mat_map,
         cond_normalizer=bundle.cond_normalizer,
         target_normalizer=bundle.target_normalizer,
+        batch_size=batch_size,
+        shuffle=False,
     )
-    _, val_ds = train_val_split(
-        full, cond_cont, cond_cat, target, val_fraction=val_fraction, seed=seed
-    )
-    return DataLoader(val_ds, batch_size=batch_size, shuffle=False)
+    return DataLoader(val_ds, batch_size=None)
 
 
 def _to_raw_targets(
@@ -316,7 +315,7 @@ def collect_samples(
     steps_kw = {} if steps is None else {"steps": steps}
 
     cond_list, real_list, gen_list = [], [], []
-    for i, (cond_cont, cond_cat, x1) in enumerate(val_loader):
+    for i, (cond_cont, cond_cat, x1, _n_sec, _sec_cont, _sec_pdg) in enumerate(val_loader):
         if n_batches is not None and i >= n_batches:
             break
         cond_cont = cond_cont.to(device)