diff --git a/README.md b/README.md
index cbf23a6..410bfea 100644
--- a/README.md
+++ b/README.md
@@ -39,12 +39,11 @@ Run:
 
 Then, inside the Docker container:
 
-```bash
-apt install -y python3-tk # This needs some configuring from your part 
+```bash 
 cd ezrknn-llm/rkllm-toolkit/examples/huggingface/
 ```
 
-Now change the `test.py` with your preferred model. This container provides Qwen-1.8B and LLaMa2 Uncensored. By default, Qwen-1.8B is selected. 
+Now change the `test.py` with your preferred model. This container provides Qwen-1.8B since it is the best working one and very lightweight. 
 Before converting the model, remember to run `git lfs pull` to download the model.
 To convert the model, run:
 
diff --git a/docker/setup.sh b/docker/setup.sh
index 0d0fb54..f726a42 100755
--- a/docker/setup.sh
+++ b/docker/setup.sh
@@ -2,9 +2,10 @@
 
 # Dependencies
 apt update
-apt install -y python3 pip git curl wget nano sudo 
+apt install -y python3 pip git curl wget nano sudo apt-utils
 
 # Better than using the default from Rockchip
+git clone https://github.com/Pelochus/ezrknn-llm/
 curl https://raw.githubusercontent.com/Pelochus/ezrknn-llm/master/install.sh | sudo bash
 
 # For running the test.py
@@ -13,12 +14,12 @@ pip install /ezrknn-llm/rkllm-toolkit/packages/rkllm_toolkit-1.0.0-cp38-cp38-lin
 # Clone some compatible LLMs
 cd /ezrknn-llm/rkllm-toolkit/examples/huggingface
 git clone https://huggingface.co/Qwen/Qwen-1_8B-Chat
-git clone https://huggingface.co/georgesung/llama2_7b_chat_uncensored
 
 # Done here to avoid cloning full repository for the Docker image
-apt install -y git-lfs # python3-tk
+apt install -y git-lfs
+
+# Needed
+DEBIAN_FRONTEND=noninteractive apt install -y python3-tk
 
 # cd Qwen-1_8B-Chat
 # git lfs pull
-# cd ../llama2_7b_chat_uncensored
-# git lfs pull