link headers with emojis.

README.md

@@ -9,12 +9,12 @@
 * [04/23] Models and datasets of SoM-LLaVA are released!
 
 ## :scroll: Contents
-- [Results](#:bar_chart:-results)
-- [Dataset](#dataset)
-- [Model Weights](#model-checkpoints)
-- [Showcases](#showcases)
-- [Training](#training)
-- [Using SoM](#notes-for-using-SoM)
+- [Results](#bar_chart-results)
+- [Dataset](#seedling-dataset)
+- [Model Weights](#cake-model-checkpoints)
+- [Showcases](#dango-showcases)
+- [Training](#mushroom-training)
+- [Using SoM](#snowflake-notes-for-using-SoM)
 
 ## :bar_chart: Results
 <table>
@@ -98,14 +98,14 @@ Two additional models for ablation study:
   <img src="examples/list1.png" width="80%"/>
 </p>
 
-## :stew: Training
+## :horse_racing: Training
 We adopt the training code of [LLaVA](https://github.com/haotian-liu/LLaVA). Please set up environments following the instructions.
 
 After downloading our data (or preparing your own SoM data), train SoM-LLaVA via command line: 
 
 `bash scripts/v1_5/finetune.sh`
 
-## Notes for using SoM
+## :snowflake: Notes for using SoM
 
 * Init virtual envs