Update README.md

README.md

@@ -33,7 +33,7 @@ model-index:
 
 # QED-Base-v3
 
-QED-Base-v3 is a **~154M parameter causal language model** pretrained from scratch by **Qarvexium**. It is a base model — it has not been instruction-tuned or aligned for chat, and is designed to continue text rather than follow instructions or hold a conversation.
+QED-Base-v3 is a **~111M parameter causal language model** pretrained from scratch by **Qarvexium**. It is a base model — it has not been instruction-tuned or aligned for chat, and is designed to continue text rather than follow instructions or hold a conversation.
 
 ## Model Details
 
@@ -50,7 +50,7 @@ QED-Base-v3 is a **~154M parameter causal language model** pretrained from scrat
 | Tokenizer             | QED-B3 tokenizer         |
 | Vocabulary size       | 56,000                   |
 | Model type            | Decoder-only Transformer |
-| Parameters            | ~154M                    |
+| Parameters            | ~111M                    |
 | Hidden size           | 768                      |
 | Layers                | 12                       |
 | Attention heads       | 12                       |