Updates Round 2

Author: Oncorporation

MANIFEST.in

@@ -8,3 +8,6 @@ include audiocraft/py.typed
 include assets/*.mp3
 include assets/*.ico
 include assets/*.png
+include assets/*.ttf
+include assets/*.svg
+include assets/*.gif

README.md

@@ -35,7 +35,7 @@ Check your video's generative metadata with https://mediaarea.net/en/MediaInfo
 Also note that I wrote an extension to Gradio for the waveform in the video after v4.48.0 removed it.
 
 The key update here is in the extend utility. We segment melody input and then condition the next segment with current tensors and tensors from the current time in the conditioning melody file.
-This allows us to make entire songs that follow the same arraigement of the original.
+This allows us to follow the same arraingement of the original melody.
 
 **Thank you Huggingface for the community grant to run this project**!!
 

app.py

@@ -132,12 +132,12 @@ def git_tag():
         except Exception:
             return "<none>"
 
-def load_melody_filepath(melody_filepath, title):
+def load_melody_filepath(melody_filepath, title, assigned_model):
     # get melody filename
     #$Union[str, os.PathLike]    
     symbols = ['_', '.', '-']
     if (melody_filepath is None) or (melody_filepath == ""):
-        return title, gr.update(maximum=0, value=0) , gr.update(value="melody-large", interactive=True)   
+        return title, gr.update(maximum=0, value=0) , gr.update(value="medium", interactive=True)
 
     if (title is None) or ("MusicGen" in title) or (title == ""):
         melody_name, melody_extension = get_filename_from_filepath(melody_filepath)
@@ -147,7 +147,10 @@ def load_melody_filepath(melody_filepath, title):
     else:
         melody_name = title
 
-    print(f"Melody name: {melody_name}, Melody Filepath: {melody_filepath}\n")
+    if ("melody" not in assigned_model):
+        assigned_model = "melody-large"
+
+    print(f"Melody name: {melody_name}, Melody Filepath: {melody_filepath}, Model: {assigned_model}\n")
 
     # get melody length in number of segments and modify the UI
     melody = get_melody(melody_filepath)
@@ -157,7 +160,7 @@ def load_melody_filepath(melody_filepath, title):
     print(f"Melody length: {len(melody_data)}, Melody segments: {total_melodys}\n")
     MAX_PROMPT_INDEX = total_melodys   
 
-    return  gr.update(value=melody_name), gr.update(maximum=MAX_PROMPT_INDEX, value=0), gr.update(value="melody", interactive=True)
+    return  gr.update(value=melody_name), gr.update(maximum=MAX_PROMPT_INDEX, value=0), gr.update(value=assigned_model, interactive=True)
 
 def predict(model, text, melody_filepath, duration, dimension, topk, topp, temperature, cfg_coef, background, title, settings_font, settings_font_color, seed, overlap=1, prompt_index = 0, include_title = True, include_settings = True, harmony_only = False):
     global MODEL, INTERRUPTED, INTERRUPTING, MOVE_TO_CPU
@@ -441,7 +444,7 @@ def ui(**kwargs):
                             text = gr.Text(label="Describe your music", interactive=True, value="4/4 100bpm 320kbps 48khz, Industrial/Electronic Soundtrack, Dark, Intense, Sci-Fi")
                             with gr.Column():                        
                                 duration = gr.Slider(minimum=1, maximum=720, value=10, label="Duration (s)", interactive=True)
-                                model = gr.Radio(["melody", "medium", "small", "large", "melody-large", "stereo-small", "stereo-medium", "stereo-large", "stereo-melody", "stereo-melody-large"], label="AI Model", value="melody", interactive=True)
+                                model = gr.Radio(["melody", "medium", "small", "large", "melody-large", "stereo-small", "stereo-medium", "stereo-large", "stereo-melody", "stereo-melody-large"], label="AI Model", value="medium", interactive=True)
                         with gr.Row():
                             submit = gr.Button("Generate", elem_id="btn-generate")
                             # Adapted from https://github.com/rkfg/audiocraft/blob/long/app.py, MIT license.
@@ -482,22 +485,21 @@ def ui(**kwargs):
                         seed_used = gr.Number(label='Seed used', value=-1, interactive=False)
 
             radio.change(toggle_audio_src, radio, [melody_filepath], queue=False, show_progress=False)
-            melody_filepath.change(load_melody_filepath, inputs=[melody_filepath, title], outputs=[title, prompt_index , model], api_name="melody_filepath_change", queue=False)
+            melody_filepath.change(load_melody_filepath, inputs=[melody_filepath, title, model], outputs=[title, prompt_index , model], api_name="melody_filepath_change", queue=False)
             reuse_seed.click(fn=lambda x: x, inputs=[seed_used], outputs=[seed], queue=False, api_name="reuse_seed")
             submit.click(predict, inputs=[model, text,melody_filepath, duration, dimension, topk, topp, temperature, cfg_coef, background, title, settings_font, settings_font_color, seed, overlap, prompt_index, include_title, include_settings, harmony_only], outputs=[output, wave_file, seed_used], api_name="submit")
             gr.Examples(
-                fn=predict,
                 examples=[
                     [
                         "4/4 120bpm 320kbps 48khz, An 80s driving pop song with heavy drums and synth pads in the background",
                         "./assets/bach.mp3",
-                        "stereo-melody-large",
+                        "melody",
                         "80s Pop Synth"
                     ],
                     [
                         "4/4 120bpm 320kbps 48khz, A cheerful country song with acoustic guitars",
                         "./assets/bolero_ravel.mp3",
-                        "melody",
+                        "stereo-melody-large",
                         "Country Guitar"
                     ],
                     [