Add 'prompt' input to image-to-image specs (#1313)

Seems like we've forgotten the most important parameter 😄 

Already used in
https://huggingface.co/stabilityai/stable-diffusion-xl-refiner-1.0
widget even though we've not documented it properly in the specs.

Author: Wauplin

packages/tasks/src/tasks/image-to-image/inference.ts

@@ -36,6 +36,10 @@ export interface ImageToImageParameters {
 	 * a higher quality image at the expense of slower inference.
 	 */
 	num_inference_steps?: number;
+	/**
+	 * The text prompt to guide the image generation.
+	 */
+	prompt?: string;
 	/**
 	 * The size in pixel of the output image.
 	 */

packages/tasks/src/tasks/image-to-image/spec/input.json

@@ -20,6 +20,10 @@
 			"title": "ImageToImageParameters",
 			"type": "object",
 			"properties": {
+				"prompt": {
+					"type": "string",
+					"description": "The text prompt to guide the image generation."
+				},
 				"guidance_scale": {
 					"type": "number",
 					"description": "For diffusion models. A higher guidance scale value encourages the model to generate images closely linked to the text prompt at the expense of lower image quality."