Update app.py

app.py

@@ -164,7 +164,7 @@ with gr.Blocks(theme=theme, css=css) as demo:
             gr.Markdown(
             f"""
             ## 🎶YourMT3+: Multi-instrument Music Transcription with Enhanced Transformer Architectures and Cross-dataset Stem Augmentation
-            ### Model card:
+            ## Model card:
             - Model name: `{model_name}`
             - Encoder backbone: Perceiver-TF + Mixture of Experts (2/8)
             - Decoder backbone: Multi-channel T5-small
@@ -173,7 +173,7 @@ with gr.Blocks(theme=theme, css=css) as demo:
             - Augmentation strategy: Intra-/Cross dataset stem augment, No Pitch-shifting
             - FP Precision: BF16-mixed for training, FP16 for inference
             
-            #### Caution:
+            ## Caution:
             - Currently running on CPU, and it takes longer than 3 minutes for a 30-second input. Please try [GPU-HuggingFace-demo](mimbres/YourMT3) for fast inference.
             - For acadmic reproduction purpose, we strongly recommend to use [Colab Demo](https://colab.research.google.com/drive/1AgOVEBfZknDkjmSRA7leoa81a2vrnhBG?usp=sharing) with multiple checkpoints.