Predict Simple Update

README.md

@@ -4,7 +4,7 @@ emoji: 🎼
 colorFrom: gray
 colorTo: red
 sdk: gradio
-sdk_version: 5.34.0
+sdk_version: 5.34.1
 python_version: 3.12.8
 app_file: app.py
 pinned: true

app.py

@@ -33,7 +33,7 @@ import librosa
 import modules.user_history
 from modules.version_info import versions_html, commit_hash, get_xformers_version
 from modules.gradio import *
-from modules.file_utils import get_file_parts, get_filename_from_filepath, convert_title_to_filename, get_unique_file_path, delete_file
+from modules.file_utils import get_file_parts, get_filename_from_filepath, convert_title_to_filename, get_unique_file_path, delete_file, download_and_save_image
 
 MODEL = None
 MODELS = None
@@ -248,7 +248,7 @@ def load_melody_filepath(melody_filepath, title, assigned_model, topp, temperatu
     
     return  gr.update(value=melody_name), gr.update(maximum=MAX_PROMPT_INDEX, value=-1), gr.update(value=assigned_model, interactive=True), gr.update(value=topp), gr.update(value=temperature), gr.update(value=cfg_coef), gr.update(maximum=MAX_OVERLAP)
 
-def predict(model, text, melody_filepath = None, duration=10, dimension=2, topk=200, topp=0, temperature=1.0, cfg_coef=4.0, background = None, title="UnlimitedMusicGen", settings_font="./assets/arial.ttf", settings_font_color = "#c87f05", seed=-1, overlap=1, prompt_index = 0, include_title = True, include_settings = True, harmony_only = False, profile = gr.OAuthProfile, segment_length = 30, settings_font_size=28, settings_animate_waveform=False, video_orientation="Landscape", excerpt_duration=3.5, progress=gr.Progress(track_tqdm=True)):
+def predict(model, text, melody_filepath = None, duration=10, dimension=2, topk=200, topp=0, temperature=1.0, cfg_coef=4.0, background = None, title="UnlimitedMusicGen", settings_font="./assets/arial.ttf", settings_font_color = "#c87f05", seed=-1, overlap=1, prompt_index = 0, include_title = True, include_settings = True, harmony_only = False, profile = gr.OAuthProfile, segment_length = 30, settings_font_size=28, settings_animate_waveform=False, video_orientation="Landscape", excerpt_duration=3.5, return_history_json=False, progress=gr.Progress(track_tqdm=True)):
     """
     Generate music and video based on the provided parameters and model.
 
@@ -278,10 +278,11 @@ def predict(model, text, melody_filepath = None, duration=10, dimension=2, topk=
         settings_animate_waveform (bool, optional): Animate waveform in video.
         video_orientation (str, optional): Video orientation.
         excerpt_duration (float, optional): Excerpt duration for style conditioning.
+        return_history_json (bool, optional): Whether to return history JSON instead of typical output. Default to False.
         progress (gr.Progress, optional): Gradio progress tracker.
 
     Returns:
-        tuple: (waveform_video_path, wave_file_path, seed_used)
+        tuple(str,str,str): (waveform_video_path, wave_file_path, seed_used)
     """
     global MODEL, INTERRUPTED, INTERRUPTING, MOVE_TO_CPU
     output_segments = None
@@ -291,9 +292,14 @@ def predict(model, text, melody_filepath = None, duration=10, dimension=2, topk=
     if melody_filepath in ["None", ""]:
         melody_filepath = None
 
-    if background in ["None", ""]:
-        background = "./assets/background.png"
+    #if background is a url string, download it using download_and_save_image
+    if background is None or background in ["None", ""]:
+        background = load_background_filepath(video_orientation)
 
+    if background.startswith("http://") or background.startswith("https://"):
+        username = profile.value.username if hasattr(profile.value, 'username') else "default_user" if (profile is None) else profile
+        background = download_and_save_image(background, modules.user_history._user_images_path(username))
+    
     if melody_filepath:
         melody_name, melody_extension = get_filename_from_filepath(melody_filepath)
         melody = get_melody(melody_filepath)
@@ -525,7 +531,7 @@ def predict(model, text, melody_filepath = None, duration=10, dimension=2, topk=
             "background": background,
             "include_title": include_title,
             "include_settings": include_settings,
-            "profile": "Satoshi Nakamoto" if profile.value is None else profile.value.username,
+            "profile": profile.value.username if hasattr(profile, 'value') and hasattr(profile.value, 'username') else "default_user" if (profile is None) else profile,
             "commit": commit_hash(),
             "tag": git_tag(),
             "version": gr.__version__,
@@ -557,8 +563,8 @@ def predict(model, text, melody_filepath = None, duration=10, dimension=2, topk=
             print(f"Error renaming file: {e}")
 
         if waveform_video_path:
-            modules.user_history.save_file(
-            profile=profile.value,
+            history_results = modules.user_history.save_file(
+            profile=profile.value.username if hasattr(profile, 'value') and hasattr(profile.value, 'username') else "default_user" if (profile is None) else profile,
             image=background,
             audio=file.name,
             video=waveform_video_path,
@@ -584,11 +590,30 @@ def predict(model, text, melody_filepath = None, duration=10, dimension=2, topk=
 
     #torch.cuda.empty_cache()
     torch.cuda.ipc_collect()
-    return waveform_video_path, file.name, seed
 
+    if  return_history_json:
+        return history_results
+    else:
+        return waveform_video_path, file.name, seed
 
+def fix_path(path: str) -> str:
+    """
+    Strips all characters preceding '_user_history' in the given path and replaces them with "./".
+    
+    If the substring '_user_history' is not found, returns the original path.
+    
+    Args:
+        path (str): The input file path.
+        
+    Returns:
+        str: The modified file path.
+    """
+    index = path.find("_user_history")
+    if index != -1:
+        return "./" + path[index:].replace("\\", "/")
+    return path
 # Add this wrapper function above the gr.api definitions
-def predict_simple(model: str, text: str, duration: int = 10, dimension: int = 2, topk: int = 200, topp: float = 0, temperature: float = 1.0, cfg_coef: float = 4.0, background: str = "./assets/background.png", title: str = "UnlimitedMusicGen", settings_font: str = "./assets/arial.ttf", settings_font_color: str = "#c87f05", seed: int = -1, overlap: int = 1, prompt_index: int = 0, include_title: bool = True, include_settings: bool = True, profile: str = "Satoshi Nakamoto", segment_length: int = 30, settings_font_size: int = 28, settings_animate_waveform: bool = False, video_orientation: str = "Landscape"):
+def predict_simple(model: str, text: str, duration: int = 10, dimension: int = 2, topk: int = 200, topp: float = 0.01, temperature: float = 1.0, cfg_coef: float = 4.0, background: str = "./assets/background.png", title: str = "UnlimitedMusicGen", settings_font: str = "./assets/arial.ttf", settings_font_color: str = "#c87f05", seed: int = -1, overlap: int = 1, prompt_index: int = -1, include_title: bool = True, include_settings: bool = True, profile: str = "Satoshi Nakamoto", segment_length: int = 30, settings_font_size: int = 28, settings_animate_waveform: bool = False, video_orientation: str = "Landscape",  return_history_json: bool = False) -> tp.List[tp.Tuple[str, str, str]]:
     """
         Generate music and video based on the provided parameters and model.
 
@@ -607,7 +632,7 @@ def predict_simple(model: str, text: str, duration: int = 10, dimension: int = 2
             settings_font_color (str, optional): Font color for settings text. Default to "
             seed (int, optional): Random seed. Default to -1.
             overlap (int, optional): Segment overlap in seconds. Default to 1.
-            prompt_index (int, optional): Melody segment index. Default to 0.
+            prompt_index (int, optional): Melody segment index. Default to -1.
             include_title (bool, optional): Whether to add title to video. Default to True.
             include_settings (bool, optional): Whether to add settings to video. Default to True.
             profile (str, optional): User profile.
@@ -615,9 +640,10 @@ def predict_simple(model: str, text: str, duration: int = 10, dimension: int = 2
             settings_font_size (int, optional): Font size for settings text.
             settings_animate_waveform (bool, optional): Animate waveform in video.
             video_orientation (str, optional): Video orientation
+            return_history_json (bool, optional): Whether to return history JSON instead of typical output. Default to False.
 
         Returns:
-            tuple: (waveform_video_path, wave_file_path, seed_used)
+             tp.List[tp.Tuple[str, str, str]]: [waveform_video_path, wave_file_path, seed_used]
     """
     profile_username_to_send = "Satoshi Nakamoto" 
 
@@ -636,7 +662,15 @@ def predict_simple(model: str, text: str, duration: int = 10, dimension: int = 2
         elif isinstance(actual_profile_data, str) and actual_profile_data: # string username
             profile_username_to_send = actual_profile_data
 
-    return predict(model, text, melody_filepath=None, duration=duration, dimension=dimension, topk=topk, topp=topp, temperature=temperature, cfg_coef=cfg_coef, background=background, title=title, settings_font=settings_font, settings_font_color=settings_font_color, seed=seed, overlap=overlap, prompt_index=prompt_index, include_title=include_title, include_settings=include_settings, harmony_only=False, profile=profile, segment_length=segment_length, settings_font_size=settings_font_size, settings_animate_waveform=settings_animate_waveform, video_orientation=video_orientation, excerpt_duration=3.5)
+    UMG_result = predict(model, text, melody_filepath=None, duration=duration, dimension=dimension, topk=topk, topp=topp, temperature=temperature, cfg_coef=cfg_coef, background=background, title=title, settings_font=settings_font, settings_font_color=settings_font_color, seed=seed, overlap=overlap, prompt_index=prompt_index, include_title=include_title, include_settings=include_settings, harmony_only=False, profile=profile, segment_length=segment_length, settings_font_size=settings_font_size, settings_animate_waveform=settings_animate_waveform, video_orientation=video_orientation, excerpt_duration=3.5,  return_history_json=return_history_json)
+    if return_history_json:
+        #content = [UMG_result["image_path"], UMG_result["video_path"], UMG_result["audio_path"], UMG_result["metadata"]["Seed"]]
+        #content = {"content":[UMG_result], "isError":False, "message":"Music generated successfully."}
+        #content = {fix_path(UMG_result["image_path"]), fix_path(UMG_result["video_path"]), fix_path(UMG_result["audio_path"]), UMG_result["metadata"]["Seed"]}        
+        content = UMG_result["video_path"], UMG_result["audio_path"], UMG_result["metadata"]["Seed"]
+        UMG_result = content
+    
+    return UMG_result
 
 gr.set_static_paths(paths=["fonts/","assets/","images/"])
 def ui(**kwargs):
@@ -792,7 +826,7 @@ def ui(**kwargs):
          ).then(
              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, user_profile, segment_length, settings_font_size, settings_animate_waveform, video_orientation],
-             outputs=[output, wave_file, seed_used], scroll_to_output=True)
+             outputs=[output, wave_file, seed_used], scroll_to_output=True, show_api=False)
 
         # Show the interface
         launch_kwargs = {}

modules/file_utils.py

@@ -1,7 +1,11 @@
 # file_utils
 import os
 import shutil
+from PIL import Image
 from pathlib import Path
+from io import BytesIO
+from urllib.parse import urlparse
+import requests
 
 def get_file_parts(file_path: str):
     # Split the path into directory and filename
@@ -115,3 +119,28 @@ def get_unique_file_path(directory, filename, file_ext, counter=0):
 
 # Example usage:
 # new_file_path = get_unique_file_path(video_dir, title_file_name, video_new_ext)
+
+def download_and_save_image(url: str, dst_folder: Path) -> Path:
+    """
+    Downloads an image from a URL, verifies it with PIL, and saves it in dst_folder with a unique filename.
+    
+    Args:
+        url (str): The image URL.
+        dst_folder (Path): The destination folder for the image.
+        
+    Returns:
+        Path: The saved image's file path.
+    """
+    response = requests.get(url)
+    response.raise_for_status()
+    pil_image = Image.open(BytesIO(response.content))
+    
+    parsed_url = urlparse(url)
+    original_filename = os.path.basename(parsed_url.path)  # e.g., "background.png"
+    base, ext = os.path.splitext(original_filename)
+    
+    # Use get_unique_file_path from file_utils.py to generate a unique file path.
+    unique_filepath_str = get_unique_file_path(str(dst_folder), base, ext)
+    dst = Path(unique_filepath_str)
+    pil_image.save(dst)
+    return dst

modules/user_history.py

@@ -18,7 +18,7 @@ Useful links:
 Update by Surn (Charles Fettinger)
 """
 
-__version__ = "0.3.6"
+__version__ = "0.3.7"
 
 import json
 import os
@@ -40,7 +40,7 @@ import wave
 from mutagen.mp3 import MP3, EasyMP3
 import torchaudio
 import subprocess
-from modules.file_utils import get_file_parts, rename_file_to_lowercase_extension
+from modules.file_utils import get_file_parts, rename_file_to_lowercase_extension, download_and_save_image
 from tqdm import tqdm
 
 user_profile = gr.State(None)
@@ -165,7 +165,7 @@ def save_image(
     # Ignore images from logged out users
     if profile is None:
         return
-    username = profile["preferred_username"]
+    username = profile if isinstance(profile, str) else profile["preferred_username"]
 
     # Ignore images if user history not used
     user_history = _UserHistory()
@@ -202,7 +202,7 @@ def save_file(
     # Ignore files from logged out users
     if profile is None:
         return
-    username = profile["preferred_username"]
+    username = profile if isinstance(profile, str) else profile["preferred_username"]
 
     # Ignore files if user history not used
     user_history = _UserHistory()
@@ -304,16 +304,17 @@ def save_file(
     # If no files were saved, nothing to do
     if image_path is None and video_path is None and audio_path is None and document_path is None:
         return
-    # else:
-    #     # Return the paths of the saved files
-    #     return {
-    #         "image_path": image_path,
-    #         "video_path": video_path,
-    #         "audio_path": audio_path,
-    #         "document_path": document_path,
-    #         "label": label,
-    #         "metadata": metadata
-    #     }, json.dumps(data)
+    else:
+         # Return the paths of the saved files
+        return {
+            "image_path": image_path,
+            "video_path": video_path,
+            "audio_path": audio_path,
+            "document_path": document_path,
+            "label": label,
+            "metadata": metadata
+        }
+        #, json.dumps(data)
     
 def get_filepath():
     """Return the path to the user history folder."""
@@ -398,7 +399,7 @@ def _fetch_user_history(profile: gr.OAuthProfile | None) -> List[Tuple[str, str]
     if profile is None:
         user_profile = gr.State(None)
         return []
-    username = str(profile["preferred_username"])
+    username = profile if isinstance(profile, str) else profile["preferred_username"]
     
     user_profile = gr.State(profile)
 
@@ -434,7 +435,7 @@ def _export_user_history(profile: gr.OAuthProfile | None) -> Dict | None:
     # Cannot load history for logged out users
     if profile is None:
         return None
-    username = profile["preferred_username"]
+    username = profile if isinstance(profile, str) else profile["preferred_username"]
 
     user_history = _UserHistory()
     if not user_history.initialized:
@@ -463,7 +464,7 @@ def _delete_user_history(profile: gr.OAuthProfile | None) -> None:
     # Cannot load history for logged out users
     if profile is None:
         return
-    username = profile["preferred_username"]
+    username = profile if isinstance(profile, str) else profile["preferred_username"]
 
     user_history = _UserHistory()
     if not user_history.initialized:
@@ -482,9 +483,14 @@ def _delete_user_history(profile: gr.OAuthProfile | None) -> None:
 def _copy_image(image: Image.Image | np.ndarray | str | Path, dst_folder: Path, uniqueId: str = "") -> Path:
     try:
         """Copy image to the images folder."""
-        # Already a path => copy it
+        # If image is a string, check if it's a URL.
         if isinstance(image, str):
-            image = Path(image)
+            if image.startswith("http://") or image.startswith("https://"):
+                return download_and_save_image(image, dst_folder)
+            else:
+                # Assume it's a local filepath string.
+                image = Path(image)
+
         if isinstance(image, Path):
             dst = dst_folder / Path(f"{uniqueId}_{Path(image).name}")  # keep file ext
             shutil.copyfile(image, dst)
@@ -645,6 +651,8 @@ def _display_if_admin() -> Callable:
             return ""
         if profile["preferred_username"] in _fetch_admins():
             return _admin_content()
+        if profile in _fetch_admins():
+            return _admin_content()
         return ""
 
     return _inner