Update train.py

train.py

@@ -10,12 +10,6 @@ import os
 import shutil
 import zipfile
 
-# Load and format your dataset
-df = pd.read_csv("data.csv")
-eos_token = tokenizer.eos_token or "</s>"  # Fallback if eos_token is None
-df["text"] = df.apply(lambda row: f"### Instruction:\n{row['instruction']}\n\n### Response:\n{row['response']} {eos_token}", axis=1)
-dataset = Dataset.from_pandas(df[["text"]])
-
 # Load Unsloth model
 model, tokenizer = FastLanguageModel.from_pretrained(
     model_name = "HuggingFaceTB/SmolLM2-1.7B",
@@ -24,6 +18,14 @@ model, tokenizer = FastLanguageModel.from_pretrained(
     load_in_4bit = True,
 )
 
+# Get eos_token after tokenizer is loaded
+eos_token = tokenizer.eos_token or "</s>"# Fallback if eos_token is None
+
+# Load and format your dataset
+df = pd.read_csv("data.csv")
+df["text"] = df.apply(lambda row: f"### Instruction:\n{row['instruction']}\n\n### Response:\n{row['response']} {eos_token}", axis=1)
+dataset = Dataset.from_pandas(df[["text"]])
+
 # Apply LoRA without task_type
 model = FastLanguageModel.get_peft_model(
     model,