diff --git a/Readme.md b/Readme.md
index 38ebc1a..a910fb5 100755
--- a/Readme.md
+++ b/Readme.md
@@ -5,6 +5,8 @@ This folder contains the code to perform Half-Quadratic Quantization (<b>HQQ</b>
 <b>HQQ</b> is a fast and accurate model quantizer that skips the need for calibration data. It's super simple to implement (just a few lines of code for the optimizer). It can crunch through quantizing the Llama2-70B model in only 4 minutes! 🚀
 
 ### Installation 
+First, make sure you have a Pytorch 2 version that matches your CUDA version: https://pytorch.org/
+
 You can install hqq via  ```pip install hqq```. 
 
 To get the latest version, you can install the core library directly via ```pip install git+https://github.com/mobiusml/hqq.git```. 
diff --git a/setup.py b/setup.py
index a63bd75..1f22332 100755
--- a/setup.py
+++ b/setup.py
@@ -9,5 +9,5 @@
     author_email='hicham@mobiuslabs.com',
     license='Apache 2',
     packages=['hqq', 'hqq/core', 'hqq/engine', 'hqq/models', 'hqq/models/hf', 'hqq/models/timm', 'hqq/models/vllm'],
-    install_requires=['numpy>=1.24.4','tqdm>=4.64.1', 'torch>=2.1.1', 'huggingface_hub', 'accelerate', 'timm', 'transformers>=4.36.1', 'termcolor'], #add vllm/langchain?
+    install_requires=['numpy>=1.24.4','tqdm>=4.64.1', 'huggingface_hub', 'accelerate', 'timm', 'transformers>=4.36.1', 'termcolor'], #'torch>=2.1.1', add vllm/langchain?
 )