updated readme and loading screens

README.md

@@ -11,3 +11,8 @@ short_description: generating drum kits using descriptive queries with CLAP
 ---
 
 Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+
+To run this demo locally, follow these steps:
+1. Clone the repository to your machine: `git clone https://huggingface.co/spaces/arhanv/drum-kit-generator`
+2. Install the dependencies with `pip install -r requirements.txt`
+3. Run the streamlit app with `streamlit run app.py`

app.py

@@ -11,6 +11,7 @@ st.title("semantic spaces: kit generator")
 st.subheader("generate drum kits and audio effects with text prompts")
 st.write("uses publicly available samples from [freesound](https://zenodo.org/records/4687854) and [CLAP embeddings](https://github.com/LAION-AI/CLAP) for text-based querying")
 st.write("hint: turn audio effects on! try weird prompts!")
+st.write("(works best on desktop, audio effects should take 20s-1min to generate)")
 
 with st.container(border=True):
     # User Inputs
@@ -25,12 +26,14 @@ with st.container(border=True):
 
     # Run the inference
     if st.button("Generate Drum Kit"):
+        st.write("Finding the best samples...")
         start_drum_time = time.time()
         drum_kit = generate_drum_kit(prompt, kit_size)
         drum_time = time.time() - start_drum_time
         st.session_state["dry_kit"] = drum_kit
-        st.write(f"Drum kit generated in {drum_time} seconds.")
+        st.write(f"Drum Kit generated in {drum_time} seconds.")
         if use_fx:
+            st.write("Generating audio effects... (takes up to 1 min)")
             start_fx_time = time.time()
             drum_kit, fx_params, pre_fx_fitness, post_fx_fitness = get_fx(drum_kit, fx_prompt)
             fx_time = time.time() - start_fx_time