Torch Object Storage
Name | Location | Type | Required | Sample value | Description |
Name | Location | Type | Required | Sample value | Description |
x-tos-process | Query | String | Yes | - | Represents the special identifier for the TextToImage API request, value is aigc/text_to_image. |
Name | Type | Required | Sample value | Description |
Name | Type | Required | Sample value | Description |
TaskType | String | Yes | text_to_image | Task type: in this scenario, the value is text_to_image, meaning the combined use of image processing features and text-to-image model. |
InputDataSource | Arrary | Yes | - | Prompt for text-to-image model. |
Prompt | String | Yes | Myriad troops | Prompt used to generate images when using the text-to-image model. Parent node: InputDataSource |
OutputProcessing | Arrary | Yes | - | After generating an image using the text-to-image model, use the object storage image processing feature to post-process the image. |
Bucket | String | Yes | tostest | The storage bucket for storing processed images. Parent node: OutputProcessing |
Object | String | Yes | save_img | Name of the processed image. Parent node: OutputProcessing |
ForbiddenOverwrite | String | Yes | false | When storing the processed image in the bucket, if a file with the same name exists in the bucket, whether to overwrite it is described as follows:
Parent node: OutputProcessing |
ProcessStyle | String | Yes | style01 | The image processing method for object storage is specified through the image style during image post-processing. For detailed information on image styles, see Creating Image Styles. Parent node: OutputProcessing |
ProcessActions | String | Yes | image/resize,w_100 | The image processing method for object storage is specified through image processing parameters during image post-processing. For detailed information on image processing parameters, see Image Processing Overview. Parent node: OutputProcessing tip Only one of the parameters ProcessStyle or ProcessActions can be configured; configuring both simultaneously is not supported. |
ModelConfig | String | Yes | - | Configure the Doubao text-to-image model. |
ModelAction | String | Yes | CVProcess | Doubao text-to-image model interface name. Parent node: ModelConfig |
ModelVersion | String | Yes | 2022-08-31 | Doubao text-to-image model interface version number. Parent node: ModelConfig |
ReqJson | String | Yes | "req_key": "high_aes_general_v21_L" | The request body parameter for the Doubao text-to-image model interface, where you set the text-to-image model configuration parameters in ReqJson to define the desired style and content of the generated images. For detailed parameter information, please refer to the interface documentation of the text-to-image model. Parent node: ModelConfig tip If both InputDataSource.Prompt and ReqJson.Prompt are configured, the final effect follows the setting of InputDataSource.Prompt. |