Text work

README.md

@@ -1,4 +1,4 @@
-# 🏆 llm-leaderboard
+# 🏆 LLM-Leaderboard
 
 A joint community effort to create one central leaderboard for LLMs. Contributions and corrections welcome!
 
@@ -73,6 +73,18 @@ We are always happy for contributions! You can contribute by the following:
     - improving the existing code
     - requesting and implementing new features
 
+## Future Ideas
+
+- add model year
+- add "export current view as .csv" button to streamlit demo
+- (TBD) add model details:
+    - #params
+    - #tokens seen during training
+    - length context window
+    - architecture type (transformer-decoder, transformer-encoder, transformer-encoder-decoder, ...)
+- if additional model details, allow to hide them in the interactive streamlit dashboard with a checkbox?
+- (TBD) improvements on the filtering in the streamlit demo, maybe filter by value range?
+
 ## More Open LLMs
 
 If you are interested in an overview about open llms for commercial use and finetuning, check out the [open-llms](https://github.com/eugeneyan/open-llms) repository.