Delete app.py

app.py

@@ -1,375 +0,0 @@
-import gradio as gr
-import numpy as np
-import random
-import torch
-import spaces
-from PIL import Image
-from diffusers import FlowMatchEulerDiscreteScheduler
-from optimization import optimize_pipeline_
-from qwenimage.pipeline_qwenimage_edit_plus import QwenImageEditPlusPipeline
-from qwenimage.transformer_qwenimage import QwenImageTransformer2DModel
-from qwenimage.qwen_fa3_processor import QwenDoubleStreamAttnProcessorFA3
-import math
-from huggingface_hub import hf_hub_download
-from safetensors.torch import load_file
-from PIL import Image
-import os
-import gradio as gr
-from gradio_client import Client, handle_file
-import tempfile
-from typing import Optional, Tuple, Any
-
-# --- Model Loading ---
-dtype = torch.bfloat16
-device = "cuda" if torch.cuda.is_available() else "cpu"
-pipe = QwenImageEditPlusPipeline.from_pretrained(
-    "Qwen/Qwen-Image-Edit-2509",
-    transformer=QwenImageTransformer2DModel.from_pretrained(
-        "linoyts/Qwen-Image-Edit-Rapid-AIO",
-        subfolder='transformer',
-        torch_dtype=dtype,
-        device_map='cuda'
-    ),
-    torch_dtype=dtype
-).to(device)
-pipe.load_lora_weights(
-    "dx8152/Qwen-Edit-2509-Multiple-angles",
-    weight_name="镜头转换.safetensors",
-    adapter_name="angles"
-)
-pipe.set_adapters(["angles"], adapter_weights=[1.])
-pipe.fuse_lora(adapter_names=["angles"], lora_scale=1.25)
-pipe.unload_lora_weights()
-pipe.transformer.__class__ = QwenImageTransformer2DModel
-pipe.transformer.set_attn_processor(QwenDoubleStreamAttnProcessorFA3())
-optimize_pipeline_(
-    pipe,
-    image=[Image.new("RGB", (1024, 1024)), Image.new("RGB", (1024, 1024))],
-    prompt="prompt"
-)
-
-MAX_SEED = np.iinfo(np.int32).max
-
-def _generate_video_segment(
-    input_image_path: str,
-    output_image_path: str,
-    prompt: str,
-    request: gr.Request,
-    duration: float = 5.0  # مدة الفيديو بالثواني
-) -> str:
-    """
-    Generate a video segment of exact duration using Wan 2.2.
-    Duration is controlled via num_frames = duration * fps.
-    """
-    x_ip_token = request.headers.get('x-ip-token', None)
-    headers = {"x-ip-token": x_ip_token} if x_ip_token else {}
-
-    video_client = Client(
-        "multimodalart/wan-2-2-first-last-frame",
-        headers=headers
-    )
-
-    fps = 25
-    num_frames = max(1, int(duration * fps))  # ضمان عدد صحيح وإيجابي
-
-    try:
-        result = video_client.predict(
-            start_image_pil=handle_file(input_image_path),
-            end_image_pil=handle_file(output_image_path),
-            prompt=prompt or "Camera movement transformation",
-            num_frames=num_frames,
-            fps=fps,
-            motion="medium",  # يمكنك تغييره إلى "high" لسلاسة أكبر
-            api_name="/generate_video",
-        )
-        return result[0]["video"]
-    except Exception as e:
-        raise gr.Error(f"فشل إنشاء الفيديو: {str(e)}")
-
-def build_camera_prompt(
-    rotate_deg: float = 0.0,
-    move_forward: float = 0.0,
-    vertical_tilt: float = 0.0,
-    wideangle: bool = False
-) -> str:
-    prompt_parts = []
-    if rotate_deg != 0:
-        direction = "left" if rotate_deg > 0 else "right"
-        abs_deg = abs(rotate_deg)
-        prompt_parts.append(
-            f"将镜头向{direction}旋转{abs_deg}度 Rotate the camera {abs_deg} degrees to the {direction}."
-        )
-    if move_forward > 5:
-        prompt_parts.append("将镜头转为特写镜头 Turn the camera to a close-up.")
-    elif move_forward >= 1:
-        prompt_parts.append("将镜头向前移动 Move the camera forward.")
-    if vertical_tilt <= -1:
-        prompt_parts.append("将相机转向鸟瞰视角 Turn the camera to a bird's-eye view.")
-    elif vertical_tilt >= 1:
-        prompt_parts.append("将相机切换到仰视视角 Turn the camera to a worm's-eye view.")
-    if wideangle:
-        prompt_parts.append(" 将镜头转为广角镜头 Turn the camera to a wide-angle lens.")
-    final_prompt = " ".join(prompt_parts).strip()
-    return final_prompt if final_prompt else "no camera movement"
-
-@spaces.GPU
-def infer_camera_edit(
-    image: Optional[Image.Image] = None,
-    rotate_deg: float = 0.0,
-    move_forward: float = 0.0,
-    vertical_tilt: float = 0.0,
-    wideangle: bool = False,
-    seed: int = 0,
-    randomize_seed: bool = True,
-    true_guidance_scale: float = 1.0,
-    num_inference_steps: int = 4,
-    height: Optional[int] = None,
-    width: Optional[int] = None,
-    prev_output: Optional[Image.Image] = None,
-) -> Tuple[Image.Image, int, str]:
-    progress = gr.Progress(track_tqdm=True)
-   
-    prompt = build_camera_prompt(rotate_deg, move_forward, vertical_tilt, wideangle)
-    print(f"Generated Prompt: {prompt}")
-
-    if randomize_seed:
-        seed = random.randint(0, MAX_SEED)
-    generator = torch.Generator(device=device).manual_seed(seed)
-
-    pil_images = []
-    if image is not None:
-        if isinstance(image, Image.Image):
-            pil_images.append(image.convert("RGB"))
-        elif hasattr(image, "name"):
-            pil_images.append(Image.open(image.name).convert("RGB"))
-    elif prev_output:
-        pil_images.append(prev_output.convert("RGB"))
-
-    if len(pil_images) == 0:
-        raise gr.Error("يرجى رفع صورة أولاً.")
-
-    if prompt == "no camera movement":
-        return pil_images[0], seed, prompt
-
-    result = pipe(
-        image=pil_images,
-        prompt=prompt,
-        height=height if height != 0 else None,
-        width=width if width != 0 else None,
-        num_inference_steps=num_inference_steps,
-        generator=generator,
-        true_cfg_scale=true_guidance_scale,
-        num_images_per_prompt=1,
-    ).images[0]
-    return result, seed, prompt
-
-def create_video_between_images(
-    input_image: Optional[Image.Image],
-    output_image: Optional[np.ndarray],
-    prompt: str,
-    request: gr.Request,
-    video_duration: float = 5.0
-) -> str:
-    if input_image is None or output_image is None:
-        raise gr.Error("كلا الصورتين مطلوبة لإنشاء الفيديو.")
-
-    try:
-        with tempfile.NamedTemporaryFile(delete=False, suffix=".png") as tmp:
-            input_image.save(tmp.name)
-            input_image_path = tmp.name
-        output_pil = Image.fromarray(output_image.astype('uint8'))
-        with tempfile.NamedTemporaryFile(delete=False, suffix=".png") as tmp:
-            output_pil.save(tmp.name)
-            output_image_path = tmp.name
-
-        video_path = _generate_video_segment(
-            input_image_path,
-            output_image_path,
-            prompt,
-            request,
-            duration=video_duration
-        )
-        # تنظيف الملفات المؤقتة
-        os.unlink(input_image_path)
-        os.unlink(output_image_path)
-        return video_path
-    except Exception as e:
-        raise gr.Error(f"فشل إنشاء الفيديو: {e}")
-
-# --- UI ---
-css = '''#col-container { max-width: 900px; margin: 0 auto; }
-.dark .progress-text{color: white !important}
-#examples{max-width: 900px; margin: 0 auto; }'''
-
-def reset_all() -> list:
-    return [0, 0, 0, False, 5.0, False, True]  # أضفنا video_duration
-
-def end_reset() -> bool:
-    return False
-
-def update_dimensions_on_upload(image: Optional[Image.Image]) -> Tuple[int, int]:
-    if image is None:
-        return 1024, 1024
-    original_width, original_height = image.size
-    if original_width > original_height:
-        new_width = 1024
-        new_height = int(new_width * (original_height / original_width))
-    else:
-        new_height = 1024
-        new_width = int(new_height * (original_width / original_height))
-    new_width = (new_width // 8) * 8
-    new_height = (new_height // 8) * 8
-    return new_width, new_height
-
-with gr.Blocks(theme=gr.themes.Citrus(), css=css) as demo:
-    with gr.Column(elem_id="col-container"):
-        gr.Markdown("## 🎬 Qwen Image Edit — Camera Angle Control + Video")
-        gr.Markdown("""
-            **تحكم كامل في زاوية الكاميرا + فيديو انتقالي بدقة 5 ثواني**  
-            باستخدام LoRA متعدد الزوايا + Wan 2.2  
-            """)
-
-        with gr.Row():
-            with gr.Column(scale=1):
-                image = gr.Image(label="الصورة الأصلية", type="pil")
-                prev_output = gr.Image(value=None, visible=False)
-                is_reset = gr.Checkbox(value=False, visible=False)
-
-                with gr.Tab("التحكم في الكاميرا"):
-                    rotate_deg = gr.Slider(label="تدوير (يمين/يسار) °", minimum=-90, maximum=90, step=45, value=0)
-                    move_forward = gr.Slider(label="تحريك للأمام → تكبير", minimum=0, maximum=10, step=5, value=0)
-                    vertical_tilt = gr.Slider(label="زاوية عمودية (طائر ↔ دودة)", minimum=-1, maximum=1, step=1, value=0)
-                    wideangle = gr.Checkbox(label="عدسة واسعة الزاوية", value=False)
-
-                with gr.Row():
-                    reset_btn = gr.Button("إعادة تعيين")
-                    run_btn = gr.Button("توليد", variant="primary")
-
-                with gr.Accordion("إعدادات متقدمة", open=False):
-                    seed = gr.Slider(label="البذرة", minimum=0, maximum=MAX_SEED, step=1, value=0)
-                    randomize_seed = gr.Checkbox(label="بذرة عشوائية", value=True)
-                    true_guidance_scale = gr.Slider(label="مقياس التوجيه", minimum=1.0, maximum=10.0, step=0.1, value=1.0)
-                    num_inference_steps = gr.Slider(label="خطوات الاستدلال", minimum=1, maximum=40, step=1, value=4)
-                    height = gr.Slider(label="الارتفاع", minimum=256, maximum=2048, step=8, value=1024)
-                    width = gr.Slider(label="العرض", minimum=256, maximum=2048, step=8, value=1024)
-
-            with gr.Column(scale=1):
-                result = gr.Image(label="الصورة المعدلة", interactive=False)
-                prompt_preview = gr.Textbox(label="النص المُرسل", interactive=False)
-
-                with gr.Accordion("إنشاء فيديو انتقالي", open=True):
-                    video_duration = gr.Slider(
-                        label="مدة الفيديو (ثواني)",
-                        minimum=1, maximum=10, step=0.5, value=5.0
-                    )
-                    create_video_button = gr.Button(
-                        "🎥 إنشاء فيديو (5 ثواني)",
-                        variant="secondary",
-                        visible=False
-                    )
-                    video_output = gr.Video(
-                        label="الفيديو الناتج",
-                        show_download_button=True,
-                        autoplay=True,
-                        height=500
-                    )
-
-    # --- Events ---
-    inputs = [
-        image, rotate_deg, move_forward, vertical_tilt, wideangle,
-        seed, randomize_seed, true_guidance_scale, num_inference_steps,
-        height, width, prev_output
-    ]
-    outputs = [result, seed, prompt_preview]
-
-    # Reset
-    reset_btn.click(
-        fn=reset_all,
-        inputs=None,
-        outputs=[rotate_deg, move_forward, vertical_tilt, wideangle, video_duration, is_reset],
-        queue=False
-    ).then(fn=end_reset, inputs=None, outputs=[is_reset], queue=False)
-
-    # Generate + show video button
-    def infer_and_show(*args):
-        result_img, result_seed, result_prompt = infer_camera_edit(*args)
-        show_button = args[0] is not None
-        return result_img, result_seed, result_prompt, gr.update(visible=show_button)
-
-    run_event = run_btn.click(
-        fn=infer_and_show,
-        inputs=inputs,
-        outputs=outputs + [create_video_button]
-    )
-
-    # Create video
-    create_video_button.click(
-        fn=lambda: gr.update(visible=True),
-        outputs=video_output,
-        api_name=False
-    ).then(
-        fn=create_video_between_images,
-        inputs=[image, result, prompt_preview, gr.Request(), video_duration],
-        outputs=video_output,
-        api_name="create_video"
-    )
-
-    # Examples
-    gr.Examples(
-        examples=[
-            ["tool_of_the_sea.png", 90, 0, 0, False, 0, True, 1.0, 4, 568, 1024],
-            ["monkey.jpg", -90, 0, 0, False, 0, True, 1.0, 4, 704, 1024],
-            ["metropolis.jpg", 0, 0, -1, False, 0, True, 1.0, 4, 816, 1024],
-            ["disaster_girl.jpg", -45, 0, 1, False, 0, True, 1.0, 4, 768, 1024],
-            ["grumpy.png", 90, 0, 1, False, 0, True, 1.0, 4, 576, 1024]
-        ],
-        inputs=[
-            image, rotate_deg, move_forward, vertical_tilt, wideangle,
-            seed, randomize_seed, true_guidance_scale, num_inference_steps, height, width
-        ],
-        outputs=outputs,
-        fn=infer_camera_edit,
-        cache_examples="lazy",
-        elem_id="examples"
-    )
-
-    # Upload → update size + reset
-    image.upload(
-        fn=update_dimensions_on_upload,
-        inputs=[image],
-        outputs=[width, height]
-    ).then(
-        fn=reset_all,
-        inputs=None,
-        outputs=[rotate_deg, move_forward, vertical_tilt, wideangle, video_duration, is_reset],
-        queue=False
-    ).then(
-        fn=end_reset,
-        inputs=None,
-        outputs=[is_reset],
-        queue=False
-    )
-
-    # Live inference
-    def maybe_infer(is_reset: bool, *args):
-        if is_reset:
-            return [gr.update()] * 4
-        result_img, result_seed, result_prompt = infer_camera_edit(*args)
-        show_button = args[0] is not None
-        return result_img, result_seed, result_prompt, gr.update(visible=show_button)
-
-    control_inputs_with_flag = [is_reset] + inputs
-    for control in [rotate_deg, move_forward, vertical_tilt, wideangle, video_duration]:
-        control.change(
-            fn=maybe_infer,
-            inputs=control_inputs_with_flag,
-            outputs=outputs + [create_video_button]
-        )
-
-    run_event.then(lambda img, *_: img, inputs=[result], outputs=[prev_output])
-
-    # API
-    gr.api(infer_camera_edit, api_name="infer_edit_camera_angles")
-    gr.api(create_video_between_images, api_name="create_video_between_images")
-
-demo.launch(mcp_server=True, show_api=True)