remove mlx dependancy

README.md

@@ -20,7 +20,6 @@ This project demonstrates a chatbot implementation using:
 - [Gradio](https://gradio.app) for the user interface
 - [Microsoft's Phi-2 model](https://huggingface.co/microsoft/phi-2) as the base language model
 - [OpenAssistant Conversations Dataset (OASST1)](https://huggingface.co/datasets/OpenAssistant/oasst1) for fine-tuning
-- GenAI code assistant 
 
 ## Fine-tuning
 
@@ -28,6 +27,10 @@ The model was fine-tuned using the QLoRA (Quantized Low-Rank Adaptation) techniq
 
 Credit for the fine-tuning process goes to the excellent guide by [Deltaaruna](https://medium.com/rahasak/fine-tune-llms-on-your-pc-with-qlora-apple-mlx-c2aedf1f607d).
 
+## Usage
+
+This app runs on CPU and uses Hugging Face's Transformers library for inference. It's designed to be compatible with various environments, including the Hugging Face Spaces platform.
+
 ## License
 
 MIT

requirements.txt

@@ -1,4 +1,5 @@
-huggingface_hub==0.22.2
-mlx>=0.0.7
-transformers
-numpy
+torch==2.0.1
+transformers==4.31.0
+gradio==4.36.1
+huggingface_hub
+accelerate