Update requirements.txt

requirements.txt

@@ -1 +1,10 @@
 gradio==4.31.0
+torch # requires torch 2.1.1+ (for efficient sdpa implementation)
+PyMuPDF # for reading PDFs with Python
+tqdm # for progress bars
+sentence-transformers # for embedding models
+accelerate # for quantization model loading
+bitsandbytes # for quantizing models (less storage space)
+flash-attn --no-build-isolation # for faster attention mechanism = faster LLM inference
+nltk
+spacy