Update backend/app/train.py

backend/app/train.py

@@ -46,7 +46,7 @@ def load_and_prepare_dataset():
 # === 2. Tokenization ===
 def tokenize_dataset(dataset, tokenizer):
     print(f"🔠 Tokenizing {len(dataset)} samples...")
-    tokenized = dataset.map(lambda x: tokenizer(x["text"], truncation=True), batched=True)
+    tokenized = dataset.map(lambda x: tokenizer(x["text"], truncation=True, max_length=128), batched=True)
     print(f"✅ Tokenized: now has keys {tokenized.column_names}")
     return tokenized