diff --git a/train_baseline.py b/train_baseline.py
index 7fb6f3d6796430d4b34f4bb396725871bf3286b9..653c326902d4ea92f457fb8ec74f825cc4b6b6b3 100644
--- a/train_baseline.py
+++ b/train_baseline.py
@@ -90,7 +90,7 @@ combinaison_label = [
 ]
 
 for ix, comb in enumerate(combinaison):
-    df = pd.concat([pd.read_csv(fn,sep="\t").head(500) for fn in comb])
+    df = pd.concat([pd.read_csv(fn,sep="\t") for fn in comb])
 
     index = NGram(n=4)
     data_vectorizer = Pipeline([