[Tasks] Add Llama-3.2-11B-Vision-Instruct as the first recommended …

…model for `image-text-to-text` (#956)

This PR is a proposition to add
`meta-llama/Llama-3.2-11B-Vision-Instruct` as the first recommended
model for
[`image-text-to-text`](https://huggingface.co/tasks/image-text-to-text)
task.
This would be nice to have, as we also want to show in the [Inference
API documentation](https://huggingface.co/docs/api-inference/index) how
to perform inference with this model, since it's both supported by TGI
and it's a popular conversational VLM.

Feel free to suggest a better description or another ranking for this
model.

cc @merveenoyan @osanseviero

packages/tasks/src/tasks/image-text-to-text/data.ts

@@ -43,8 +43,8 @@ const taskData: TaskDataCustom = {
 	metrics: [],
 	models: [
 		{
-			description: "Cutting-edge vision language model that can take multiple image inputs.",
-			id: "facebook/chameleon-7b",
+			description: "Powerful vision language model with great visual understanding and reasoning capabilities.",
+			id: "meta-llama/Llama-3.2-11B-Vision-Instruct",
 		},
 		{
 			description: "Cutting-edge conversational vision language model that can take multiple image inputs.",