fix: multi-turn data breaks

data.py

@@ -15,8 +15,8 @@ def tokenize_inputs(config, tokenizer, examples):
     out = {"labels": [], "input_ids": []}
     for prompt, response in zip(examples["prompt"], examples["response"]):
         if different_eos:
-            if response.count("</s>") > 0:
-                response = response.replace("</s>", tokenizer.eos_token)
+            if response.count("</s> \n") > 0:
+                response = response.replace("</s> \n", f"{tokenizer.eos_token} \n") 
 
         prompt_len = len(tokenizer(prompt + "\n", return_tensors="pt")["input_ids"][0])