Update app.py

app.py

@@ -1,3 +1,6 @@
+# ------------------------------------------------------------
+#  IMPORTS
+# ------------------------------------------------------------
 import spaces
 import torch
 import requests
@@ -19,7 +22,7 @@ from torchao.quantization import Int8WeightOnlyConfig
 import aoti
 
 # ------------------------------------------------------------
-# -------------------------- CONFIG ---------------------------
+#  CONFIG
 # ------------------------------------------------------------
 MODEL_ID = "Wan-AI/Wan2.2-I2V-A14B-Diffusers"
 
@@ -45,7 +48,7 @@ default_negative_prompt = (
 )
 
 # ------------------------------------------------------------
-# ----------------------- MODEL LOADING -----------------------
+#  MODEL LOADING
 # ------------------------------------------------------------
 pipe = WanImageToVideoPipeline.from_pretrained(
     MODEL_ID,
@@ -70,6 +73,7 @@ pipe.load_lora_weights(
     weight_name="Lightx2v/lightx2v_I2V_14B_480p_cfg_step_distill_rank128_bf16.safetensors",
     adapter_name="lightx2v",
 )
+
 kwargs_lora = {"load_into_transformer_2": True}
 pipe.load_lora_weights(
     "Kijai/WanVideo_comfy",
@@ -77,6 +81,7 @@ pipe.load_lora_weights(
     adapter_name="lightx2v_2",
     **kwargs_lora,
 )
+
 pipe.set_adapters(["lightx2v", "lightx2v_2"], adapter_weights=[1.0, 1.0])
 pipe.fuse_lora(adapter_names=["lightx2v"], lora_scale=3.0, components=["transformer"])
 pipe.fuse_lora(adapter_names=["lightx2v_2"], lora_scale=1.0, components=["transformer_2"])
@@ -91,7 +96,7 @@ aoti.aoti_blocks_load(pipe.transformer, "zerogpu-aoti/Wan2", variant="fp8da")
 aoti.aoti_blocks_load(pipe.transformer_2, "zerogpu-aoti/Wan2", variant="fp8da")
 
 # ------------------------------------------------------------
-# -------------------------- HELPERS -------------------------
+#  HELPERS
 # ------------------------------------------------------------
 def resize_image(image: Image.Image) -> Image.Image:
     """Resize / crop the input image so the model receives a valid size."""
@@ -117,6 +122,7 @@ def resize_image(image: Image.Image) -> Image.Image:
         top = (height - crop_h) // 2
         img = image.crop((0, top, width, top + crop_h))
     else:
+        # No cropping needed – just compute target size
         if width > height:  # landscape
             target_w = MAX_DIM
             target_h = int(round(target_w / aspect_ratio))
@@ -125,6 +131,7 @@ def resize_image(image: Image.Image) -> Image.Image:
             target_w = int(round(target_h * aspect_ratio))
         img = image
 
+    # Round to the nearest multiple of MULTIPLE_OF and clamp
     final_w = round(target_w / MULTIPLE_OF) * MULTIPLE_OF
     final_h = round(target_h / MULTIPLE_OF) * MULTIPLE_OF
     final_w = max(MIN_DIM, min(MAX_DIM, final_w))
@@ -134,6 +141,7 @@ def resize_image(image: Image.Image) -> Image.Image:
 
 
 def get_num_frames(duration_seconds: float) -> int:
+    """Number of frames the model will generate for the requested duration."""
     return 1 + int(
         np.clip(
             int(round(duration_seconds * FIXED_FPS)),
@@ -153,9 +161,12 @@ def get_duration(
     guidance_scale_2,
     seed,
     randomize_seed,
-    progress,
+    progress,            # <-- required by @spaces.GPU
 ):
-    """Estimate how long the GPU will be needed – used by @spaces.GPU."""
+    """
+    Rough estimate of how long the GPU will be occupied.
+    Used by the @spaces.GPU decorator to enforce the 30‑second safety cap.
+    """
     BASE_FRAMES_HEIGHT_WIDTH = 81 * 832 * 624
     BASE_STEP_DURATION = 15
 
@@ -165,34 +176,10 @@ def get_duration(
     step_duration = BASE_STEP_DURATION * factor ** 1.5
     est = 10 + int(steps) * step_duration
 
-    # safety cap – we never want to block the GPU >30 s
+    # Never block the GPU > 30 s
     return min(est, 30)
 
 
-@spaces.GPU
-def translate_albanian_to_english(text):
-    """Optional helper – not used in the UI but kept unchanged."""
-    if not text.strip():
-        raise gr.Error("Please enter a description.")
-    for attempt in range(2):
-        try:
-            resp = requests.post(
-                "https://hal1993-mdftranslation1234567890abcdef1234567890-fc073a6.hf.space/v1/translate",
-                json={"from_language": "sq", "to_language": "en", "input_text": text},
-                headers={"accept": "application/json", "Content-Type": "application/json"},
-                timeout=5,
-            )
-            resp.raise_for_status()
-            return resp.json().get("translate", "")
-        except Exception as e:
-            if attempt == 1:
-                raise gr.Error("Translation failed. Please try again.") from e
-    raise gr.Error("Translation failed. Please try again.")
-
-
-# ------------------------------------------------------------
-# -------------------------- MAIN FUNCTION ---------------------
-# ------------------------------------------------------------
 @spaces.GPU(duration=get_duration)
 def generate_video(
     input_image,
@@ -204,9 +191,12 @@ def generate_video(
     guidance_scale_2=1.5,
     seed=42,
     randomize_seed=False,
-    progress=None,                 # ← made optional – no UI change
+    progress=gr.Progress(track_tqdm=True),   # <-- now mandatory
 ):
-    """Generate a video from an image + prompt."""
+    """
+    Generate a video from an image + prompt.
+    Returns (video_path, seed_used).
+    """
     if input_image is None:
         raise gr.Error("Please upload an input image.")
 
@@ -247,12 +237,12 @@ def generate_video(
 
 
 # ------------------------------------------------------------
-# --------------------------- UI -------------------------------
+#  UI – unchanged visual / CSS / 500‑guard / unique‑link
 # ------------------------------------------------------------
 def create_demo():
     with gr.Blocks(css="", title="Fast Image to Video") as demo:
         # -----------------------------------------------------------------
-        # 500‑error guard – **exactly the same** as in your original file
+        # 500‑error guard – exactly the same as in your fork
         # -----------------------------------------------------------------
         gr.HTML(
             """
@@ -266,7 +256,7 @@ def create_demo():
         )
 
         # -----------------------------------------------------------------
-        # All your custom CSS / visual theme – **unaltered**
+        # Custom CSS – kept verbatim
         # -----------------------------------------------------------------
         gr.HTML(
             """
@@ -276,7 +266,6 @@ def create_demo():
             @keyframes glow-hover {0%{box-shadow:0 0 20px rgba(0,255,128,0.7);}50%{box-shadow:0 0 20px rgba(0,255,128,0.9);}100%{box-shadow:0 0 20px rgba(0,255,128,0.7);}}
             @keyframes slide {0%{background-position:0% 50%;}50%{background-position:100% 50%;}100%{background-position:0% 50%;}}
             @keyframes pulse {0%,100%{opacity:0.7;}50%{opacity:1;}}
-            @keyframes typewriter {0%{width:0;}100%{width:100%;}}
             body{
                 background:#000 !important;
                 color:#FFF !important;
@@ -295,7 +284,7 @@ def create_demo():
             body::before{
                 content:"";
                 display:block;
-                height:600px;               /* <-- this is the top gap you asked for */
+                height:600px;               /* <-- top gap you asked for */
                 background:#000 !important;
             }
             .gr-blocks,.container{
@@ -378,7 +367,7 @@ def create_demo():
                 box-sizing:border-box !important;
                 display:block !important;
             }
-            /* FORCE HIDE ALL GRADIO PROCESSING ELEMENTS - 100+ SELECTORS */
+            /* HIDE ALL GRADIO PROCESSING UI – 100+ SELECTORS */
             .image-container[aria-label="Generated Video"] .progress-text,
             .image-container[aria-label="Generated Video"] .gr-progress,
             .image-container[aria-label="Generated Video"] .gr-progress-bar,
@@ -433,23 +422,9 @@ def create_demo():
             .image-container[aria-label="Input Image"] .file-upload,
             .image-container[aria-label="Input Image"] .file-preview,
             .image-container[aria-label="Input Image"] .image-actions,
-            .image-container[aria-label="Input Image"] .gr-file-upload,
-            .image-container[aria-label="Input Image"] .gr-file,
-            .image-container[aria-label="Input Image"] .gr-actions,
-            .image-container[aria-label="Input Image"] .gr-upload-button,
-            .image-container[aria-label="Input Image"] .gr-image-toolbar,
-            .image-container[aria-label="Input Image"] .gr-file-actions,
-            .image-container[aria-label="Input Image"] .gr-upload-options,
             .image-container[aria-label="Generated Video"] .file-upload,
             .image-container[aria-label="Generated Video"] .file-preview,
-            .image-container[aria-label="Generated Video"] .image-actions,
-            .image-container[aria-label="Generated Video"] .gr-file-upload,
-            .image-container[aria-label="Generated Video"] .gr-file,
-            .image-container[aria-label="Generated Video"] .gr-actions,
-            .image-container[aria-label="Generated Video"] .gr-upload-button,
-            .image-container[aria-label="Generated Video"] .gr-image-toolbar,
-            .image-container[aria-label="Generated Video"] .gr-file-actions,
-            .image-container[aria-label="Generated Video"] .gr-upload-options{
+            .image-container[aria-label="Generated Video"] .image-actions{
                 display:none!important;
             }
             .image-container[aria-label="Generated Video"].processing{
@@ -483,10 +458,6 @@ def create_demo():
             .image-container[aria-label="Generated Video"].processing *{
                 display:none!important;
             }
-            .image-container[aria-label="Generated Video"].processing video,
-            .image-container[aria-label="Generated Video"].processing img{
-                display:none!important;
-            }
             input,textarea,.gr-dropdown,.gr-dropdown select{
                 background:#000!important;
                 color:#FFF!important;
@@ -497,10 +468,6 @@ def create_demo():
                 max-width:100vw!important;
                 box-sizing:border-box!important;
             }
-            input:hover,textarea:hover,.gr-dropdown:hover,.gr-dropdown select:hover{
-                box-shadow:0 0 8px rgba(255,255,255,0.3)!important;
-                transition:box-shadow .3s;
-            }
             .gr-button-primary{
                 background:linear-gradient(90deg,rgba(0,255,128,0.3),rgba(0,200,100,0.3),rgba(0,255,128,0.3))!important;
                 background-size:200% 100%;
@@ -561,10 +528,7 @@ def create_demo():
                 .gr-button-primary:hover{
                     box-shadow:0 0 12px rgba(0,255,128,0.9)!important;
                 }
-                .image-container{
-                    min-height:300px;
-                    box-shadow:0 0 8px rgba(255,255,255,0.3)!important;
-                }
+                .image-container{min-height:300px;}
                 .image-container[aria-label="Generated Video"].processing::before{
                     font-size:1.2rem!important;
                 }
@@ -574,7 +538,7 @@ def create_demo():
         )
 
         # -----------------------------------------------------------------
-        # UI layout – **exactly the same structure you built**
+        # UI layout – unchanged component order (matches generate_video signature)
         # -----------------------------------------------------------------
         with gr.Row(elem_id="general_items"):
             gr.Markdown("# ")
@@ -614,7 +578,7 @@ def create_demo():
                 )
 
         # -----------------------------------------------------------------
-        # Wiring – unchanged component order (matches generate_video signature)
+        # Wiring – keep the same order as the function signature
         # -----------------------------------------------------------------
         generate_button.click(
             fn=generate_video,
@@ -628,6 +592,7 @@ def create_demo():
                 gr.State(value=1.5),                    # guidance_scale_2
                 gr.State(value=42),                     # seed
                 gr.State(value=True),                   # randomize_seed
+                # progress is *not* passed – the @spaces.GPU decorator injects it
             ],
             outputs=[output_video, gr.State(value=42)],
         )
@@ -635,6 +600,9 @@ def create_demo():
     return demo
 
 
+# ------------------------------------------------------------
+#  MAIN
+# ------------------------------------------------------------
 if __name__ == "__main__":
     demo = create_demo()
     # keep the launch flags you originally used