Update app.py

app.py

@@ -27,6 +27,7 @@ tokenizer, generator_conf, model = load_model(REPO_NAME)
 total_params = sum(p.numel() for p in model.parameters())
 trainable_params = sum(p.numel() for p in model.parameters() if p.requires_grad)
 embed_params = sum(p.numel() for p in model.model.embed_tokens.parameters())*2
+non_embed_params = (trainable_params - embed_params) / 1e6
 
 st.markdown(f"*This chat uses the {REPO_NAME} model with {model.get_memory_footprint() / 1e6:.2f} MB memory footprint. ")
 
@@ -34,7 +35,7 @@ st.markdown(f"*This chat uses the {REPO_NAME} model with {model.get_memory_footp
 # st.markdown(f"Total number of trainable parameters: {trainable_params}. ")
 # st.markdown(f"Total number of embed parameters: {embed_params}. ")
 
-st.markdown(f"Total number of non embedding trainable parameters: {(trainable_params - embed_params) / 1e6:.2f } million. ")
+st.markdown(f"Total number of non embedding trainable parameters: {non_embed_params:.2f} million. ")
 st.markdown(f"You may ask questions such as 'What is biology?' or 'What is the human body?'*")
 
 try: