processor: 传入apply_chat_template等处理文本的工具

processor: 传入apply_chat_template等处理文本的工具

processor: 传入apply_chat_template等处理文本的工具

processor: 传入apply_chat_template等处理文本的工具

Initialize the OpenAI client and settings. :param api_url: Base URL of the VLM API endpoint. :param key_name_of_api_key: Env

Initialize LiteLLM serving instance. Args: model: Model name (e.g., "gpt-4o", "claude-3-sonnet", "command-r-plus

:param api_url: Base URL of the cloud API (e.g. "https://api.openai.com/v1") :param image_io: ImageIO instance, for saving generated

参数： - template: 带命名占位符的字符串，例如： \"\"\"Descriptions: {descriptions} Collect all details for {type}

Initialize the VideoFilteredClipGenerator operator with default parameters.

Initialize VideoVQAGenerator with API model parameters.

Initialize the long video cotqa pipeline with default API model configurations.

Initialize VideoCOTQATest with API model parameters.

Initialize the VideoSceneClipGenerator operator with default parameters.

Initialize VideoVQAGenerator with default parameters.