Issues with loading models in 8bit in Colab

Hello. I am using Google Colab. I run this code
!pip install sentencepiece -q
!pip install accelerate -q
!pip install bitsandbytes -q
from transformers import AutoTokenizer, AutoModelForCausalLM
tk = AutoTokenizer.from_pretrained(“microsoft/Orca-2-13b”)
model =AutoModelForCausalLM.from_pretrained(“microsoft/Orca-2-13b”, load_in_8bit=True)

but I always get error
ImportError: Using load_in_8bit=True requires Accelerate: pip install accelerate and the latest version of bitsandbytes pip install -i https://test.pypi.org/simple/ bitsandbytes or pip install bitsandbytes` .
They are already installed. Why this happens?