Upload 10 files

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+assets/scene_1.png filter=lfs diff=lfs merge=lfs -text

.gitignore

@@ -0,0 +1,41 @@
+__pycache__/
+*.pyc
+*.pyo
+*.pyd
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+
+# Logs
+logs/*.log
+*.log
+
+# Environment variables
+.env
+.venv
+env/
+venv/
+
+# IDE
+.vscode/
+.idea/
+*.swp
+*.swo
+*~
+
+# OS
+.DS_Store
+Thumbs.db

app.py

@@ -0,0 +1,275 @@
+# main.py
+# 主入口文件，负责启动 Gradio UI
+import gradio as gr
+from config import SCENE_CONFIGS, MODEL_CHOICES, MODE_CHOICES
+from backend_api import submit_to_backend, get_task_status, get_task_result
+from logging_utils import log_access, log_submission, is_request_allowed
+from simulation import stream_simulation_results, convert_to_h264
+from ui_components import update_history_display, update_scene_display, update_log_display, get_scene_instruction
+import os
+from datetime import datetime
+
+SESSION_TASKS = {}
+
+def run_simulation(scene, model, mode, prompt, history, request: gr.Request):
+    timestamp = datetime.now().strftime("%Y-%m-%d %H:%M:%S")
+    scene_desc = SCENE_CONFIGS.get(scene, {}).get("description", scene)
+    user_ip = request.client.host if request else "unknown"
+    session_id = request.session_hash
+    if not is_request_allowed(user_ip):
+        log_submission(scene, prompt, model, user_ip, "IP blocked temporarily")
+        raise gr.Error("Too many requests from this IP. Please wait and try again one minute later.")
+    # 传递model和mode给后端
+    #submission_result = submit_to_backend(scene, prompt, user=model)  # 可根据后端接口调整
+    submission_result = submit_to_backend(scene, prompt, mode, model, user_ip)
+    if submission_result.get("status") != "pending":
+        log_submission(scene, prompt, model, user_ip, "Submission failed")
+        raise gr.Error(f"Submission failed: {submission_result.get('message', 'unknown issue')}")
+    try:
+        task_id = submission_result["task_id"]
+        SESSION_TASKS[session_id] = task_id
+        gr.Info(f"Simulation started, task_id: {task_id}")
+        import time
+        time.sleep(5)
+        status = get_task_status(task_id)
+        result_folder = status.get("result", "")
+    except Exception as e:
+        log_submission(scene, prompt, model, user_ip, str(e))
+        raise gr.Error(f"error occurred when parsing submission result from backend: {str(e)}")
+    if not os.path.exists(result_folder):
+        log_submission(scene, prompt, model, user_ip, "Result folder provided by backend doesn't exist")
+        raise gr.Error(f"Result folder provided by backend doesn't exist: <PATH>{result_folder}")
+    try:
+        for video_path in stream_simulation_results(result_folder, task_id):
+            if video_path:
+                yield video_path, history
+    except Exception as e:
+        log_submission(scene, prompt, model, user_ip, str(e))
+        raise gr.Error(f"流式输出过程中出错: {str(e)}")
+    status = get_task_status(task_id)
+    if status.get("status") == "completed":
+        video_path = os.path.join(status.get("result"), "output.mp4")
+        video_path = convert_to_h264(video_path)
+        new_entry = {
+            "timestamp": timestamp,
+            "scene": scene,
+            "model": model,
+            "mode": mode,
+            "prompt": prompt,
+            "video_path": video_path
+        }
+        updated_history = history + [new_entry]
+        if len(updated_history) > 10:
+            updated_history = updated_history[:10]
+        log_submission(scene, prompt, model, user_ip, "success")
+        gr.Info("Simulation completed successfully!")
+        yield None, updated_history
+    elif status.get("status") == "failed":
+        log_submission(scene, prompt, model, user_ip, status.get('result', 'backend error'))
+        raise gr.Error(f"任务执行失败: {status.get('result', 'backend 未知错误')}")
+        yield None, history
+    elif status.get("status") == "terminated":
+        log_submission(scene, prompt, model, user_ip, "terminated")
+        video_path = os.path.join(result_folder, "output.mp4")
+        if os.path.exists(video_path):
+            return f"⚠️ 任务 {task_id} 被终止，已生成部分结果", video_path, history
+        else:
+            return f"⚠️ 任务 {task_id} 被终止，未生成结果", None, history
+    else:
+        log_submission(scene, prompt, model, user_ip, "missing task's status from backend")
+        raise gr.Error("missing task's status from backend")
+        yield None, history
+
+def cleanup_session(request: gr.Request):
+    session_id = request.session_hash
+    task_id = SESSION_TASKS.pop(session_id, None)
+    from config import BACKEND_URL
+    import requests
+    if task_id:
+        try:
+            requests.post(f"{BACKEND_URL}/predict/terminate/{task_id}", timeout=3)
+        except Exception:
+            pass
+
+def record_access(request: gr.Request):
+    user_ip = request.client.host if request else "unknown"
+    user_agent = request.headers.get("user-agent", "unknown")
+    log_access(user_ip, user_agent)
+    return update_log_display()
+
+
+
+custom_css = """
+#simulation-panel {
+    border-radius: 8px;
+    padding: 20px;
+    background: #f9f9f9;
+    box-shadow: 0 2px 4px rgba(0,0,0,0.1);
+}
+#result-panel {
+    border-radius: 8px;
+    padding: 20px;
+    background: #f0f8ff;
+}
+.dark #simulation-panel { background: #2a2a2a; }
+.dark #result-panel { background: #1a2a3a; }
+.history-container {
+    max-height: 600px;
+    overflow-y: auto;
+    margin-top: 20px;
+}
+.history-accordion {
+    margin-bottom: 10px;
+}
+"""
+
+header_html = """
+<div style="display: flex; justify-content: space-between; align-items: center; width: 100%; margin-bottom: 20px; padding: 20px; background: linear-gradient(135deg, #e0e5ec 0%, #a7b5d0 100%); border-radius: 8px; box-shadow: 0 2px 8px rgba(0,0,0,0.1);">
+    <div style="display: flex; align-items: center;">
+        <img src="https://www.shlab.org.cn/static/img/index_14.685f6559.png" alt="Institution Logo" style="height: 60px; margin-right: 20px;">
+        <div>
+            <h1 style="margin: 0; color: #2c3e50; font-weight: 600;">🤖 InternManip Model Inference Demo</h1>
+            <p style="margin: 4px 0 0 0; color: #5d6d7e; font-size: 0.9em;">Model trained on InternManip framework</p>
+        </div>
+    </div>
+    <div style="display: flex; gap: 15px; align-items: center;">
+        <a href="https://github.com/OpenRobotLab" target="_blank" style="text-decoration: none; transition: transform 0.2s;" onmouseover="this.style.transform='scale(1.1)'" onmouseout="this.style.transform='scale(1)'">
+            <img src="https://github.githubassets.com/images/modules/logos_page/GitHub-Mark.png" alt="GitHub" style="height: 30px;">
+        </a>
+        <a href="https://huggingface.co/OpenRobotLab" target="_blank" style="text-decoration: none; transition: transform 0.2s;" onmouseover="this.style.transform='scale(1.1)'" onmouseout="this.style.transform='scale(1)'">
+            <img src="https://huggingface.co/front/assets/huggingface_logo-noborder.svg" alt="HuggingFace" style="height: 30px;">
+        </a>
+        <a href="http://123.57.187.96:55004/" target="_blank">
+            <button style="padding: 8px 15px; background: #3498db; color: white; border: none; border-radius: 4px; cursor: pointer; font-weight: 500; transition: all 0.2s;" 
+                    onmouseover="this.style.backgroundColor='#2980b9'; this.style.transform='scale(1.05)'" 
+                    onmouseout="this.style.backgroundColor='#3498db'; this.style.transform='scale(1)'">
+                Go to InternManip Demo
+            </button>
+        </a>
+    </div>
+</div>
+"""
+
+
+
+with gr.Blocks(title="InternNav Model Inference Demo", css=custom_css) as demo:
+    gr.HTML(header_html)
+    
+    history_state = gr.State([])
+    with gr.Row():
+        with gr.Column(elem_id="simulation-panel"):
+            gr.Markdown("### Simulation Settings")
+            scene_dropdown = gr.Dropdown(
+                label="Choose a scene",
+                choices=list(SCENE_CONFIGS.keys()),
+                value="demo1",
+                interactive=True
+            )
+            scene_description = gr.Markdown("")
+            scene_preview = gr.Image(
+                label="Scene Preview",
+                elem_classes=["scene-preview"],
+                interactive=False
+            )
+            prompt_input = gr.Textbox(
+                label="Navigation Prompt",
+                value="Walk past the left side of the bed and stop in the doorway.",
+                placeholder="e.g.: 'Walk past the left side of the bed and stop in the doorway.'",
+                lines=2,
+                max_lines=4
+            )
+            model_dropdown = gr.Dropdown(
+                label="Chose a pretrained model",
+                choices=MODEL_CHOICES,
+                value=MODEL_CHOICES[0],
+                interactive=True
+            )
+            mode_dropdown = gr.Dropdown(
+                label="Select Mode",
+                choices=MODE_CHOICES,
+                value=MODE_CHOICES[0],
+                interactive=True
+            )
+            scene_dropdown.change(
+                fn=lambda scene: [update_scene_display(scene)[0], update_scene_display(scene)[1], get_scene_instruction(scene)],
+                inputs=scene_dropdown,
+                outputs=[scene_description, scene_preview, prompt_input]
+            )
+            
+            submit_btn = gr.Button("Start Navigation Simulation", variant="primary")
+        with gr.Column(elem_id="result-panel"):
+            gr.Markdown("### Latest Simulation Result")
+            video_output = gr.Video(
+                label="Live",
+                interactive=False,
+                format="mp4",
+                autoplay=True,
+                streaming=True
+            )
+            with gr.Column() as history_container:
+                gr.Markdown("### History")
+                gr.Markdown("#### History will be reset after refresh")
+                history_slots = []
+                for i in range(10):
+                    with gr.Column(visible=False) as slot:
+                        with gr.Accordion(visible=False, open=False) as accordion:
+                            video = gr.Video(interactive=False)
+                            detail_md = gr.Markdown()
+                    history_slots.append((slot, accordion, video, detail_md))
+    with gr.Accordion("查看系统访问日志(DEV ONLY)", open=False):
+        logs_display = gr.Markdown()
+        refresh_logs_btn = gr.Button("刷新日志", variant="secondary")
+        refresh_logs_btn.click(
+            update_log_display,
+            outputs=logs_display
+        )
+    gr.Examples(
+        examples=[
+            ["demo1", "rdp", "vlnPE", "Walk past the left side of the bed and stop in the doorway."],
+            ["demo2", "rdp", "vlnPE", "Walk through the bathroom, past the sink and toilet. Stop in front of the counter with the two suitcase."],
+            ["demo3", "rdp", "vlnPE", "Do a U-turn. Walk forward through the kitchen, heading to the black door. Walk out of the door and take a right onto the deck. Walk out on to the deck and stop."],
+            ["demo4", "rdp", "vlnPE", "Walk out of bathroom and stand on white bath mat."],
+            ["demo5", "rdp", "vlnPE", "Walk straight through the double wood doors, follow the red carpet straight to the next doorway and stop where the carpet splits off."]
+        ],
+        inputs=[scene_dropdown, model_dropdown, mode_dropdown, prompt_input],
+        label="Navigation Task Examples"
+    )
+    submit_btn.click(
+        fn=run_simulation,
+        inputs=[scene_dropdown, model_dropdown, mode_dropdown, prompt_input, history_state],
+        outputs=[video_output, history_state],
+        queue=True,
+        api_name="run_simulation"
+    ).then(
+        fn=update_history_display,
+        inputs=history_state,
+        outputs=[comp for slot in history_slots for comp in slot],
+        queue=True
+    ).then(
+        fn=update_log_display,
+        outputs=logs_display,
+    )
+    demo.load(
+        fn=lambda: update_scene_display("demo1"),
+        outputs=[scene_description, scene_preview]
+    ).then(
+        fn=update_log_display,
+        outputs=logs_display
+    )
+    demo.load(
+        fn=record_access,
+        inputs=None,
+        outputs=logs_display,
+        queue=False
+    )
+    demo.queue(default_concurrency_limit=8)
+    demo.unload(fn=cleanup_session)
+
+if __name__ == "__main__":
+    demo.launch(
+        server_name="0.0.0.0", 
+        server_port=7860,  # Hugging Face Space默认端口
+        share=False, 
+        debug=False,  # 生产环境建议关闭debug
+        allowed_paths=["./assets", "./logs"]  # 修改为相对路径
+    )

app_demo.py

@@ -0,0 +1,275 @@
+# main.py
+# 主入口文件，负责启动 Gradio UI
+import gradio as gr
+from config import SCENE_CONFIGS, MODEL_CHOICES, MODE_CHOICES
+from backend_api import submit_to_backend, get_task_status, get_task_result
+from logging_utils import log_access, log_submission, is_request_allowed
+from simulation import stream_simulation_results, convert_to_h264
+from ui_components import update_history_display, update_scene_display, update_log_display, get_scene_instruction
+import os
+from datetime import datetime
+
+SESSION_TASKS = {}
+
+def run_simulation(scene, model, mode, prompt, history, request: gr.Request):
+    timestamp = datetime.now().strftime("%Y-%m-%d %H:%M:%S")
+    scene_desc = SCENE_CONFIGS.get(scene, {}).get("description", scene)
+    user_ip = request.client.host if request else "unknown"
+    session_id = request.session_hash
+    if not is_request_allowed(user_ip):
+        log_submission(scene, prompt, model, user_ip, "IP blocked temporarily")
+        raise gr.Error("Too many requests from this IP. Please wait and try again one minute later.")
+    # 传递model和mode给后端
+    #submission_result = submit_to_backend(scene, prompt, user=model)  # 可根据后端接口调整
+    submission_result = submit_to_backend(scene, prompt, mode, model, user_ip)
+    if submission_result.get("status") != "pending":
+        log_submission(scene, prompt, model, user_ip, "Submission failed")
+        raise gr.Error(f"Submission failed: {submission_result.get('message', 'unknown issue')}")
+    try:
+        task_id = submission_result["task_id"]
+        SESSION_TASKS[session_id] = task_id
+        gr.Info(f"Simulation started, task_id: {task_id}")
+        import time
+        time.sleep(5)
+        status = get_task_status(task_id)
+        result_folder = status.get("result", "")
+    except Exception as e:
+        log_submission(scene, prompt, model, user_ip, str(e))
+        raise gr.Error(f"error occurred when parsing submission result from backend: {str(e)}")
+    if not os.path.exists(result_folder):
+        log_submission(scene, prompt, model, user_ip, "Result folder provided by backend doesn't exist")
+        raise gr.Error(f"Result folder provided by backend doesn't exist: <PATH>{result_folder}")
+    try:
+        for video_path in stream_simulation_results(result_folder, task_id):
+            if video_path:
+                yield video_path, history
+    except Exception as e:
+        log_submission(scene, prompt, model, user_ip, str(e))
+        raise gr.Error(f"流式输出过程中出错: {str(e)}")
+    status = get_task_status(task_id)
+    if status.get("status") == "completed":
+        video_path = os.path.join(status.get("result"), "output.mp4")
+        video_path = convert_to_h264(video_path)
+        new_entry = {
+            "timestamp": timestamp,
+            "scene": scene,
+            "model": model,
+            "mode": mode,
+            "prompt": prompt,
+            "video_path": video_path
+        }
+        updated_history = history + [new_entry]
+        if len(updated_history) > 10:
+            updated_history = updated_history[:10]
+        log_submission(scene, prompt, model, user_ip, "success")
+        gr.Info("Simulation completed successfully!")
+        yield None, updated_history
+    elif status.get("status") == "failed":
+        log_submission(scene, prompt, model, user_ip, status.get('result', 'backend error'))
+        raise gr.Error(f"任务执行失败: {status.get('result', 'backend 未知错误')}")
+        yield None, history
+    elif status.get("status") == "terminated":
+        log_submission(scene, prompt, model, user_ip, "terminated")
+        video_path = os.path.join(result_folder, "output.mp4")
+        if os.path.exists(video_path):
+            return f"⚠️ 任务 {task_id} 被终止，已生成部分结果", video_path, history
+        else:
+            return f"⚠️ 任务 {task_id} 被终止，未生成结果", None, history
+    else:
+        log_submission(scene, prompt, model, user_ip, "missing task's status from backend")
+        raise gr.Error("missing task's status from backend")
+        yield None, history
+
+def cleanup_session(request: gr.Request):
+    session_id = request.session_hash
+    task_id = SESSION_TASKS.pop(session_id, None)
+    from config import BACKEND_URL
+    import requests
+    if task_id:
+        try:
+            requests.post(f"{BACKEND_URL}/predict/terminate/{task_id}", timeout=3)
+        except Exception:
+            pass
+
+def record_access(request: gr.Request):
+    user_ip = request.client.host if request else "unknown"
+    user_agent = request.headers.get("user-agent", "unknown")
+    log_access(user_ip, user_agent)
+    return update_log_display()
+
+
+
+custom_css = """
+#simulation-panel {
+    border-radius: 8px;
+    padding: 20px;
+    background: #f9f9f9;
+    box-shadow: 0 2px 4px rgba(0,0,0,0.1);
+}
+#result-panel {
+    border-radius: 8px;
+    padding: 20px;
+    background: #f0f8ff;
+}
+.dark #simulation-panel { background: #2a2a2a; }
+.dark #result-panel { background: #1a2a3a; }
+.history-container {
+    max-height: 600px;
+    overflow-y: auto;
+    margin-top: 20px;
+}
+.history-accordion {
+    margin-bottom: 10px;
+}
+"""
+
+header_html = """
+<div style="display: flex; justify-content: space-between; align-items: center; width: 100%; margin-bottom: 20px; padding: 20px; background: linear-gradient(135deg, #e0e5ec 0%, #a7b5d0 100%); border-radius: 8px; box-shadow: 0 2px 8px rgba(0,0,0,0.1);">
+    <div style="display: flex; align-items: center;">
+        <img src="https://www.shlab.org.cn/static/img/index_14.685f6559.png" alt="Institution Logo" style="height: 60px; margin-right: 20px;">
+        <div>
+            <h1 style="margin: 0; color: #2c3e50; font-weight: 600;">🤖 InternManip Model Inference Demo</h1>
+            <p style="margin: 4px 0 0 0; color: #5d6d7e; font-size: 0.9em;">Model trained on InternManip framework</p>
+        </div>
+    </div>
+    <div style="display: flex; gap: 15px; align-items: center;">
+        <a href="https://github.com/OpenRobotLab" target="_blank" style="text-decoration: none; transition: transform 0.2s;" onmouseover="this.style.transform='scale(1.1)'" onmouseout="this.style.transform='scale(1)'">
+            <img src="https://github.githubassets.com/images/modules/logos_page/GitHub-Mark.png" alt="GitHub" style="height: 30px;">
+        </a>
+        <a href="https://huggingface.co/OpenRobotLab" target="_blank" style="text-decoration: none; transition: transform 0.2s;" onmouseover="this.style.transform='scale(1.1)'" onmouseout="this.style.transform='scale(1)'">
+            <img src="https://huggingface.co/front/assets/huggingface_logo-noborder.svg" alt="HuggingFace" style="height: 30px;">
+        </a>
+        <a href="http://123.57.187.96:55004/" target="_blank">
+            <button style="padding: 8px 15px; background: #3498db; color: white; border: none; border-radius: 4px; cursor: pointer; font-weight: 500; transition: all 0.2s;" 
+                    onmouseover="this.style.backgroundColor='#2980b9'; this.style.transform='scale(1.05)'" 
+                    onmouseout="this.style.backgroundColor='#3498db'; this.style.transform='scale(1)'">
+                Go to InternManip Demo
+            </button>
+        </a>
+    </div>
+</div>
+"""
+
+
+
+with gr.Blocks(title="InternNav Model Inference Demo", css=custom_css) as demo:
+    gr.HTML(header_html)
+    
+    history_state = gr.State([])
+    with gr.Row():
+        with gr.Column(elem_id="simulation-panel"):
+            gr.Markdown("### Simulation Settings")
+            scene_dropdown = gr.Dropdown(
+                label="Choose a scene",
+                choices=list(SCENE_CONFIGS.keys()),
+                value="demo1",
+                interactive=True
+            )
+            scene_description = gr.Markdown("")
+            scene_preview = gr.Image(
+                label="Scene Preview",
+                elem_classes=["scene-preview"],
+                interactive=False
+            )
+            prompt_input = gr.Textbox(
+                label="Navigation Prompt",
+                value="Walk past the left side of the bed and stop in the doorway.",
+                placeholder="e.g.: 'Walk past the left side of the bed and stop in the doorway.'",
+                lines=2,
+                max_lines=4
+            )
+            model_dropdown = gr.Dropdown(
+                label="Chose a pretrained model",
+                choices=MODEL_CHOICES,
+                value=MODEL_CHOICES[0],
+                interactive=True
+            )
+            mode_dropdown = gr.Dropdown(
+                label="Select Mode",
+                choices=MODE_CHOICES,
+                value=MODE_CHOICES[0],
+                interactive=True
+            )
+            scene_dropdown.change(
+                fn=lambda scene: [update_scene_display(scene)[0], update_scene_display(scene)[1], get_scene_instruction(scene)],
+                inputs=scene_dropdown,
+                outputs=[scene_description, scene_preview, prompt_input]
+            )
+            
+            submit_btn = gr.Button("Start Navigation Simulation", variant="primary")
+        with gr.Column(elem_id="result-panel"):
+            gr.Markdown("### Latest Simulation Result")
+            video_output = gr.Video(
+                label="Live",
+                interactive=False,
+                format="mp4",
+                autoplay=True,
+                streaming=True
+            )
+            with gr.Column() as history_container:
+                gr.Markdown("### History")
+                gr.Markdown("#### History will be reset after refresh")
+                history_slots = []
+                for i in range(10):
+                    with gr.Column(visible=False) as slot:
+                        with gr.Accordion(visible=False, open=False) as accordion:
+                            video = gr.Video(interactive=False)
+                            detail_md = gr.Markdown()
+                    history_slots.append((slot, accordion, video, detail_md))
+    with gr.Accordion("查看系统访问日志(DEV ONLY)", open=False):
+        logs_display = gr.Markdown()
+        refresh_logs_btn = gr.Button("刷新日志", variant="secondary")
+        refresh_logs_btn.click(
+            update_log_display,
+            outputs=logs_display
+        )
+    gr.Examples(
+        examples=[
+            ["demo1", "rdp", "vlnPE", "Walk past the left side of the bed and stop in the doorway."],
+            ["demo2", "rdp", "vlnPE", "Walk through the bathroom, past the sink and toilet. Stop in front of the counter with the two suitcase."],
+            ["demo3", "rdp", "vlnPE", "Do a U-turn. Walk forward through the kitchen, heading to the black door. Walk out of the door and take a right onto the deck. Walk out on to the deck and stop."],
+            ["demo4", "rdp", "vlnPE", "Walk out of bathroom and stand on white bath mat."],
+            ["demo5", "rdp", "vlnPE", "Walk straight through the double wood doors, follow the red carpet straight to the next doorway and stop where the carpet splits off."]
+        ],
+        inputs=[scene_dropdown, model_dropdown, mode_dropdown, prompt_input],
+        label="Navigation Task Examples"
+    )
+    submit_btn.click(
+        fn=run_simulation,
+        inputs=[scene_dropdown, model_dropdown, mode_dropdown, prompt_input, history_state],
+        outputs=[video_output, history_state],
+        queue=True,
+        api_name="run_simulation"
+    ).then(
+        fn=update_history_display,
+        inputs=history_state,
+        outputs=[comp for slot in history_slots for comp in slot],
+        queue=True
+    ).then(
+        fn=update_log_display,
+        outputs=logs_display,
+    )
+    demo.load(
+        fn=lambda: update_scene_display("demo1"),
+        outputs=[scene_description, scene_preview]
+    ).then(
+        fn=update_log_display,
+        outputs=logs_display
+    )
+    demo.load(
+        fn=record_access,
+        inputs=None,
+        outputs=logs_display,
+        queue=False
+    )
+    demo.queue(default_concurrency_limit=8)
+    demo.unload(fn=cleanup_session)
+
+if __name__ == "__main__":
+    demo.launch(
+        server_name="0.0.0.0", 
+        server_port=7860,  # Hugging Face Space默认端口
+        share=False, 
+        debug=False,  # 生产环境建议关闭debug
+        allowed_paths=["./assets", "./logs"]  # 修改为相对路径
+    )

backend_api.py

@@ -0,0 +1,53 @@
+# backend_api.py
+# 后端API交互相关
+import requests
+import uuid
+import json
+from typing import Optional
+from config import API_ENDPOINTS
+
+def submit_to_backend(scene: str, prompt: str, mode: str, model_type: str, user: str = "Gradio-user") -> dict:
+    job_id = str(uuid.uuid4())
+    data = {
+        "model_type": model_type,
+        "instruction": prompt,
+        "episode_type": scene,
+        "mode": mode,
+    }
+    payload = {
+        "user": user,
+        "task": "robot_navigation",
+        "job_id": job_id,
+        "data": json.dumps(data)
+    }
+    try:
+        headers = {"Content-Type": "application/json"}
+        response = requests.post(
+            API_ENDPOINTS["submit_task"],
+            json=payload,
+            headers=headers,
+            timeout=200
+        )
+        return response.json()
+    except Exception as e:
+        return {"status": "error", "message": str(e)}
+
+def get_task_status(task_id: str) -> dict:
+    try:
+        response = requests.get(f"{API_ENDPOINTS['query_status']}/{task_id}", timeout=5)
+        try:
+            return response.json()
+        except json.JSONDecodeError:
+            return {"status": "error", "message": response.text}
+    except Exception as e:
+        return {"status": "error", "message": str(e)}
+
+def get_task_result(task_id: str) -> Optional[dict]:
+    try:
+        response = requests.get(
+            f"{API_ENDPOINTS['get_result']}/{task_id}",
+            timeout=5
+        )
+        return response.json()
+    except Exception as e:
+        return None

config.py

@@ -0,0 +1,46 @@
+# config.py
+# 配置相关：API、场景等
+import os
+
+BACKEND_URL = os.getenv("BACKEND_URL", "http://localhost:8000")
+API_ENDPOINTS = {
+    "submit_task": f"{BACKEND_URL}/predict/video",
+    "query_status": f"{BACKEND_URL}/predict/task",
+    "get_result": f"{BACKEND_URL}//predict"
+}
+
+SCENE_CONFIGS = {
+    "demo1": {
+        "description": "Demo 1",
+        "objects": ["bedroom", "kitchen", "living room", ""],
+        "preview_image": "./assets/scene_1.png",
+        "default_instruction": "Walk past the left side of the bed and stop in the doorway."
+    },
+    "demo2": {
+        "description": "Demo 2",
+        "objects": ["office", "meeting room", "corridor"],
+        "preview_image": "./assets/scene_2.png",
+        "default_instruction": "Walk through the bathroom, past the sink and toilet. Stop in front of the counter with the two suitcase."
+    },
+    "demo3": {
+        "description": "Demo 3",
+        "objects": ["garage", "workshop", "storage"],
+        "preview_image": "./assets/scene_3.png",
+        "default_instruction": "Do a U-turn. Walk forward through the kitchen, heading to the black door. Walk out of the door and take a right onto the deck. Walk out on to the deck and stop."
+    },
+    "demo4": {
+        "description": "Demo 4",
+        "objects": ["garden", "patio", "pool"],
+        "preview_image": "./assets/scene_4.png",
+        "default_instruction": "Walk out of bathroom and stand on white bath mat."
+    },
+    "demo5": {
+        "description": "Demo 5",
+        "objects": ["library", "hall", "lounge"],
+        "preview_image": "./assets/scene_5.png",
+        "default_instruction": "Walk straight through the double wood doors, follow the red carpet straight to the next doorway and stop where the carpet splits off."
+    },
+}
+
+MODEL_CHOICES = ["rdp", "cma"]
+MODE_CHOICES = ["vlnPE", "vlnCE"]

logging_utils.py

@@ -0,0 +1,89 @@
+# logging_utils.py
+# 日志相关工具
+import os
+import json
+from datetime import datetime,timedelta
+from collections import defaultdict
+
+LOG_DIR = "/opt/nav-fronted/logs"
+ACCESS_LOG = os.path.join(LOG_DIR, "access.log")
+SUBMISSION_LOG = os.path.join(LOG_DIR, "submissions.log")
+
+os.makedirs(LOG_DIR, exist_ok=True)
+
+IP_REQUEST_RECORDS = defaultdict(list)
+IP_LIMIT = 5
+
+def is_request_allowed(ip: str) -> bool:
+    now = datetime.now()
+    IP_REQUEST_RECORDS[ip] = [t for t in IP_REQUEST_RECORDS[ip] if now - t < timedelta(minutes=1)]
+    if len(IP_REQUEST_RECORDS[ip]) < IP_LIMIT:
+        IP_REQUEST_RECORDS[ip].append(now)
+        return True
+    return False
+
+def log_access(user_ip: str = None, user_agent: str = None):
+    timestamp = datetime.now().strftime("%Y-%m-%d %H:%M:%S")
+    log_entry = {
+        "timestamp": timestamp,
+        "type": "access",
+        "user_ip": user_ip or "unknown",
+        "user_agent": user_agent or "unknown"
+    }
+    with open(ACCESS_LOG, "a") as f:
+        f.write(json.dumps(log_entry) + "\n")
+
+def log_submission(scene: str, prompt: str, model: str, user: str = "anonymous", res: str = "unknown"):
+    timestamp = datetime.now().strftime("%Y-%m-%d %H:%M:%S")
+    log_entry = {
+        "timestamp": timestamp,
+        "type": "submission",
+        "user": user,
+        "scene": scene,
+        "prompt": prompt,
+        "model": model,
+        "res": res
+    }
+    with open(SUBMISSION_LOG, "a") as f:
+        f.write(json.dumps(log_entry) + "\n")
+
+def read_logs(log_type: str = "all", max_entries: int = 50) -> list:
+    logs = []
+    if log_type in ["all", "access"]:
+        try:
+            with open(ACCESS_LOG, "r") as f:
+                for line in f:
+                    logs.append(json.loads(line.strip()))
+        except FileNotFoundError:
+            pass
+    if log_type in ["all", "submission"]:
+        try:
+            with open(SUBMISSION_LOG, "r") as f:
+                for line in f:
+                    logs.append(json.loads(line.strip()))
+        except FileNotFoundError:
+            pass
+    logs.sort(key=lambda x: x["timestamp"], reverse=True)
+    return logs[:max_entries]
+
+def format_logs_for_display(logs: list) -> str:
+    if not logs:
+        return "No log record"
+    markdown = "### System Access Log\n\n"
+    markdown += "| Time | Type | User/IP | Details |\n"
+    markdown += "|------|------|---------|----------|\n"
+    for log in logs:
+        timestamp = log.get("timestamp", "unknown")
+        log_type = "Access" if log.get("type") == "access" else "Submission"
+        if log_type == "Access":
+            user = log.get("user_ip", "unknown")
+            details = f"User-Agent: {log.get('user_agent', 'unknown')}"
+        else:
+            user = log.get("user", "anonymous")
+            result = log.get('res', 'unknown')
+            if result != "success":
+                if len(result) > 40:
+                    result = f"{result[:20]}...{result[-20:]}"
+            details = f"Scene: {log.get('scene', 'unknown')}, Prompt: {log.get('prompt', '')}, Model: {log.get('model', 'unknown')}, result: {result}"
+        markdown += f"| {timestamp} | {log_type} | {user} | {details} |\n"
+    return markdown

requirements.txt

@@ -0,0 +1,4 @@
+gradio>=4.0.0
+requests>=2.28.0
+opencv-python>=4.6.0
+numpy>=1.21.0

simulation.py

@@ -0,0 +1,115 @@
+# simulation.py
+# 仿真与视频相关
+import os
+import time
+import uuid
+import cv2
+import numpy as np
+from typing import List
+import gradio as gr
+from backend_api import get_task_status
+
+def stream_simulation_results(result_folder: str, task_id: str, fps: int = 6):
+    result_folder = os.path.join(result_folder, "images")
+    os.makedirs(result_folder, exist_ok=True)
+    frame_buffer: List[np.ndarray] = []
+    frames_per_segment = fps * 2
+    processed_files = set()
+    width, height = 0, 0
+    last_status_check = 0
+    status_check_interval = 5
+    max_time = 240
+    while max_time > 0:
+        max_time -= 1
+        current_time = time.time()
+        if current_time - last_status_check > status_check_interval:
+            status = get_task_status(task_id)
+            if status.get("status") == "completed":
+                process_remaining_images(result_folder, processed_files, frame_buffer)
+                if frame_buffer:
+                    yield create_video_segment(frame_buffer, fps, width, height)
+                break
+            elif status.get("status") == "failed":
+                raise gr.Error(f"任务执行失败: {status.get('result', '未知错误')}")
+            elif status.get("status") == "terminated":
+                break
+            last_status_check = current_time
+        current_files = sorted(
+            [f for f in os.listdir(result_folder) if f.lower().endswith(('.png', '.jpg', '.jpeg'))],
+            key=lambda x: os.path.splitext(x)[0]
+        )
+        new_files = [f for f in current_files if f not in processed_files]
+        has_new_frames = False
+        for filename in new_files:
+            try:
+                img_path = os.path.join(result_folder, filename)
+                frame = cv2.imread(img_path)
+                if frame is not None:
+                    if width == 0:
+                        height, width = frame.shape[:2]
+                    frame_buffer.append(frame)
+                    processed_files.add(filename)
+                    has_new_frames = True
+            except Exception:
+                pass
+        if has_new_frames and len(frame_buffer) >= frames_per_segment:
+            segment_frames = frame_buffer[:frames_per_segment]
+            frame_buffer = frame_buffer[frames_per_segment:]
+            yield create_video_segment(segment_frames, fps, width, height)
+        time.sleep(1)
+    if max_time <= 0:
+        raise gr.Error("timeout 240s")
+
+def create_video_segment(frames: List[np.ndarray], fps: int, width: int, height: int) -> str:
+    os.makedirs("/opt/gradio_demo/tasks/video_chunk", exist_ok=True)
+    segment_name = f"/opt/gradio_demo/tasks/video_chunk/output_{uuid.uuid4()}.mp4"
+    fourcc = cv2.VideoWriter_fourcc(*'mp4v')
+    out = cv2.VideoWriter(segment_name, fourcc, fps, (width, height))
+    for frame in frames:
+        out.write(frame)
+    out.release()
+    return segment_name
+
+def process_remaining_images(result_folder: str, processed_files: set, frame_buffer: List[np.ndarray]):
+    current_files = sorted(
+        [f for f in os.listdir(result_folder) if f.lower().endswith(('.png', '.jpg', '.jpeg'))],
+        key=lambda x: os.path.splitext(x)[0]
+    )
+    new_files = [f for f in current_files if f not in processed_files]
+    for filename in new_files:
+        try:
+            img_path = os.path.join(result_folder, filename)
+            frame = cv2.imread(img_path)
+            if frame is not None:
+                frame_buffer.append(frame)
+                processed_files.add(filename)
+        except Exception:
+            pass
+
+def convert_to_h264(video_path):
+    import shutil
+    base, ext = os.path.splitext(video_path)
+    video_path_h264 = f"{base}_h264.mp4"
+    ffmpeg_bin = "/root/anaconda3/envs/gradio/bin/ffmpeg"
+    if not os.path.exists(ffmpeg_bin):
+        ffmpeg_bin = shutil.which("ffmpeg")
+    if ffmpeg_bin is None:
+        raise RuntimeError("❌ 找不到 ffmpeg，请确保其已安装并在 PATH 中")
+    ffmpeg_cmd = [
+        ffmpeg_bin,
+        "-i", video_path,
+        "-c:v", "libx264",
+        "-preset", "slow",
+        "-crf", "23",
+        "-c:a", "aac",
+        "-movflags", "+faststart",
+        video_path_h264
+    ]
+    import subprocess
+    try:
+        result = subprocess.run(ffmpeg_cmd, check=True, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
+        if not os.path.exists(video_path_h264):
+            raise FileNotFoundError(f"⚠️ H.264 文件未生成: {video_path_h264}")
+        return video_path_h264
+    except Exception as e:
+        raise

ui_components.py

@@ -0,0 +1,43 @@
+# ui_components.py
+# Gradio界面相关和辅助函数
+import gradio as gr
+from config import SCENE_CONFIGS
+from logging_utils import read_logs, format_logs_for_display
+
+def update_history_display(history: list) -> list:
+    updates = []
+    for i in range(10):
+        if i < len(history):
+            entry = history[i]
+            label_text = f"Simulation {i+1}  scene: {entry['scene']}, model: {entry.get('model','')}, mode: {entry.get('mode','')}, prompt: {entry['prompt']}"
+            updates.extend([
+                gr.update(visible=True),
+                gr.update(visible=True, label=label_text, open=False),
+                gr.update(value=entry['video_path'], visible=True),
+                gr.update(value=f"{entry['timestamp']}")
+            ])
+        else:
+            updates.extend([
+                gr.update(visible=False),
+                gr.update(visible=False),
+                gr.update(value=None, visible=False),
+                gr.update(value="")
+            ])
+    return updates
+
+def update_scene_display(scene: str):
+    config = SCENE_CONFIGS.get(scene, {})
+    desc = config.get("description", "No Description")
+    objects = "、".join(config.get("objects", []))
+    image = config.get("preview_image", None)
+    markdown = f"**{desc}**  \nPlaces Included: {objects}"
+    return markdown, image
+
+def get_scene_instruction(scene: str):
+    """根据场景获取默认指令"""
+    config = SCENE_CONFIGS.get(scene, {})
+    return config.get("default_instruction", "")
+
+def update_log_display():
+    logs = read_logs()
+    return format_logs_for_display(logs)