updating info

README.md

@@ -1,5 +1,5 @@
 ---
-title: Zero Gpu Slot Machine
+title: stable-melodyflow
 emoji: 🏢
 colorFrom: gray
 colorTo: blue

app.py

@@ -343,6 +343,43 @@ with gr.Blocks(title="🎵 Stable Audio Loop Generator") as iface:
     gr.Markdown("# 🎵 Stable Audio Loop Generator")
     gr.Markdown("**Generate synchronized drum and instrument loops with stable-audio-open-small, then transform with MelodyFlow!**")
     
+    # ========== MODELS & PROJECT INFO ==========
+    with gr.Accordion("📚 About the Models & Project", open=False):
+        
+        with gr.Accordion("🚀 stable-audio-open-small", open=False):
+            gr.Markdown("""
+            **stable-audio-open-small** is an incredibly fast model from the zachs and friends at Stability AI. It's capable of generating 12 seconds of audio in under a second, which gives rise to a lot of very interesting kinds of UX.
+            
+            **Note about generation speed in this zerogpu space:** You'll notice generation times are a little slower here than if you were to use the model on a local GPU. That's just a result of the way zerogpu spaces work I think... let me know if there's a way to keep the model loaded in a zerogpu space!
+            
+            **Links:**
+            - 🤗 [Model on HuggingFace](https://huggingface.co/stabilityai/stable-audio-open-small)
+            - 🐳 [Docker API Implementation](https://github.com/betweentwomidnights/stable-audio-api)
+            """)
+        
+        with gr.Accordion("🎛️ MelodyFlow", open=False):
+            gr.Markdown("""
+            **MelodyFlow** is a model by Meta that can use regularized latent inversion to do transformations of input audio.
+            
+            It's not officially a part of the audiocraft repo yet, but we use it as a docker container in the backend for gary4live.
+            
+            **Links:**
+            - 🤗 [MelodyFlow Space](https://huggingface.co/spaces/Facebook/MelodyFlow)  
+            - 🐳 [Standalone API Implementation](https://github.com/betweentwomidnights/melodyflow)
+            """)
+        
+        with gr.Accordion("🎹 gary4live Project", open=False):
+            gr.Markdown("""
+            **gary4live** is a free/open source project that uses these models, along with MusicGen, inside of Ableton Live. I run a backend myself so that we can all experiment with it, but you can also spin the backend up locally using docker-compose with our repo.
+            
+            **Project Links:**
+            - 🎵 [Main Project Repo](https://github.com/betweentwomidnights/gary4live)
+            - 🖥️ [Backend Implementation](https://github.com/betweentwomidnights/gary-backend-combined)
+            
+            **Installers:**
+            - 💿 [PC & Mac Installers on Gumroad](https://thepatch.gumroad.com/l/gary4live)
+            """)
+    
     with gr.Accordion("How This Works", open=False):
         gr.Markdown("""
         **Workflow:**