Description of the bug:

The final result of fine-tuned model responds to the prompt is,
Answer:

The median of the list [5, 2, 9, 1, 7, 4, 6, 3, 8] is 4. 

<end_of_turn>

but the correct median should be 5.

Actual vs expected behavior:

No response

Any other information you'd like to share?

No response