opt

agent/video_preprocessor/webpage_vp.py

@@ -42,7 +42,7 @@ class WebpageVideoPreprocessor(BaseLLMBackend, BaseWorker):
     kernel_size: Optional[int] = None
     show_progress: bool = True
     
-    use_cache: bool = False
+    use_cache: bool = True
     cache_dir: str = "./video_cache"
     
     @field_validator("stt", mode="before")
@@ -87,7 +87,7 @@ class WebpageVideoPreprocessor(BaseLLMBackend, BaseWorker):
         
         cache_path = (
             Path(self.cache_dir)
-            .joinpath(video_path.replace("/", "-"))
+            .joinpath(Path(video_path).name.replace("/", "-"))
             .joinpath("video_cache.pkl")
         )
         # Load video from cache if available
@@ -174,7 +174,8 @@ class WebpageVideoPreprocessor(BaseLLMBackend, BaseWorker):
                 pickle.dump(video.scenes, f)
         
         # Process video if not loaded from cache
-        if not self.stm(self.workflow_instance_id).get("video", None):
+        # if not self.stm(self.workflow_instance_id).get("video", None):
+        if not cache_path.exists():
             video = VideoScenes.load(
                 video_path=video_path,
                 threshold=self.scene_detect_threshold,
@@ -241,10 +242,10 @@ class WebpageVideoPreprocessor(BaseLLMBackend, BaseWorker):
                     self.callback.error(f"Failed to process scene {index}: {e}")
                     scene.summary = None
         
-        # if self.use_cache and not cache_path.exists():
-        #     cache_path.parent.mkdir(parents=True, exist_ok=True)
-        #     with open(cache_path, "wb") as f:
-        #         pickle.dump(video.scenes, f)
+        if self.use_cache and not cache_path.exists():
+            cache_path.parent.mkdir(parents=True, exist_ok=True)
+            with open(cache_path, "wb") as f:
+                pickle.dump(video.scenes, f)
         return {
             "video_md5": video_md5,
             "video_path": video_path,

video_cache/Build a Video Search and Summarization Agent blueprint by nvidia _ NVIDIA NIM - - 2024-11-15 18-35-47.mp4/video_cache.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:321ab99f6c22dfb8cabe2a69a733c9377ef5d5f37664e9ed7f05a15f16bac04c
+size 2901

video_cache/Build a Video Search and Summarization Agent blueprint by nvidia _ NVIDIA NIM - - 2024-11-15 18-36-01.mp4/video_cache.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bf531dcac5d926b1a4b9d5fa6ed69cb2f051894ab1582cdfc5de0ed40d85d250
+size 4447

video_cache/Build a Video Search and Summarization Agent blueprint by nvidia _ NVIDIA NIM.mp4/video_cache.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d8b552fa7034b8d6fa7a85b410d971f6d1fa0fc1b1e3162a48aae05ebc7ed7f8
+size 1514

webpage_process_video_understanding.json

@@ -0,0 +1,23 @@
+{
+    "name": "webpage_process_video_understanding",
+    "tasks": [
+        {
+            "name": "WebpageVideoPreprocessor",
+            "taskReferenceName": "webpage_video_preprocess",
+            "inputParameters": {
+                "video_path": "${workflow.input.video_path}"
+            },
+            "type": "SIMPLE",
+            "taskDefinition": {}
+        }
+    ],
+    "inputParameters": [],
+    "outputParameters": {},
+    "failureWorkflow": "",
+    "schemaVersion": 2,
+    "workflowStatusListenerEnabled": false,
+    "ownerEmail": "[email protected]",
+    "timeoutSeconds": 60,
+    "variables": {},
+    "inputTemplate": {}
+}

webpage_video_understanding.json

@@ -0,0 +1,122 @@
+{
+    "name": "webpage_video_understanding",
+    "tasks": [
+        {
+            "name": "WebpageVideoPreprocessor",
+            "taskReferenceName": "webpage_video_preprocess",
+            "inputParameters": {
+                "video_path": "${workflow.input.video_path}"
+            },
+            "type": "SIMPLE",
+            "taskDefinition": {}
+        },
+        {
+            "name": "WebpageVideoQA",
+            "taskReferenceName": "webpage_video_qa",
+            "inputParameters": {
+                "video_md5": "${workflow.input.video_md5}",
+                "video_path": "${workflow.input.video_path}",
+                "instance_id": "${workflow.input.instance_id}",
+                "question": "${workflow.input.question}"
+            },
+            "type": "SIMPLE",
+            "taskDefinition": {}
+        },
+        {
+            "name": "ConstructDncPayload",
+            "taskReferenceName": "construct_dnc_payload",
+            "inputParameters": {
+                "query": "${webpage_video_qa.output.query}"
+            },
+            "type": "SIMPLE",
+            "taskDefinition": {}
+        },
+        {
+            "name": "dncloop_task",
+            "taskReferenceName": "dncloop_task",
+            "inputParameters": {},
+            "type": "DO_WHILE",
+            "taskDefinition": {},
+            "loopCondition": " if ( $.task_exit_monitor['exit_flag'] == true) { false; } else { true; }",
+            "loopOver": [
+                {
+                    "name": "StructureUpdate",
+                    "taskReferenceName": "structure_update",
+                    "inputParameters": {
+                        "dnc_structure": "${construct_dnc_payload.output.dnc_structure}"
+                    },
+                    "type": "SIMPLE"
+                },
+                {
+                    "name": "TaskConqueror",
+                    "taskReferenceName": "task_conqueror",
+                    "inputParameters": {
+                        "dnc_structure": "${structure_update.output.dnc_structure}",
+                        "last_output": "${structure_update.output.last_output}"
+                    },
+                    "type": "SIMPLE"
+                },
+                {
+                    "name": "switch_task",
+                    "taskReferenceName": "switch_task",
+                    "inputParameters": {
+                        "switchCaseValue": "${task_conqueror.output.switch_case_value}"
+                    },
+                    "type": "SWITCH",
+                    "decisionCases": {
+                        "complex": [
+                            {
+                                "name": "TaskDivider",
+                                "taskReferenceName": "task_divider",
+                                "inputParameters": {
+                                    "dnc_structure": "${task_conqueror.output.dnc_structure}",
+                                    "last_output": "${task_conqueror.output.last_output}"
+                                },
+                                "type": "SIMPLE"
+                            }
+                        ],
+                        "failed": [
+                            {
+                                "name": "TaskRescue",
+                                "taskReferenceName": "task_rescue",
+                                "inputParameters": {
+                                    "dnc_structure": "${task_conqueror.output.dnc_structure}",
+                                    "last_output": "${task_conqueror.output.last_output}"
+                                },
+                                "type": "SIMPLE"
+                            }
+                        ]
+                    },
+                    "defaultCase": [],
+                    "evaluatorType": "value-param",
+                    "expression": "switchCaseValue"
+                },
+                {
+                    "name": "TaskExitMonitor",
+                    "taskReferenceName": "task_exit_monitor",
+                    "inputParameters": {},
+                    "type": "SIMPLE"
+                }
+            ]
+        },
+        {
+            "name": "WebpageConclude",
+            "taskReferenceName": "webpage_task_conclude",
+            "inputParameters": {
+                "dnc_structure": "${task_exit_monitor.output.dnc_structure}",
+                "last_output": "${task_exit_monitor.output.last_output}"
+            },
+            "type": "SIMPLE",
+            "taskDefinition": {}
+        }
+    ],
+    "inputParameters": [],
+    "outputParameters": {},
+    "failureWorkflow": "",
+    "schemaVersion": 2,
+    "workflowStatusListenerEnabled": false,
+    "ownerEmail": "[email protected]",
+    "timeoutSeconds": 60,
+    "variables": {},
+    "inputTemplate": {}
+}