[Inference API] Add image-text-to-text task and fix generate script

[hanouticelina]

This PR does two things:

• adds image-text-to-text task in the documentation.
• adds conversational VLMs in Chat Completion documentation.
• upgrades huggingface/tasks and generate the documentations.

[HuggingFaceDocBuilderDev]

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.

[hanouticelina]

The PR documentation is not deployed correctly and returns 404 🤔 did I miss something?

[Wauplin]

Thanks for working on this @hanouticelina! Seems like you uncovered some annoying discrepancies 🙈

The PR documentation is not deployed correctly and returns 404 🤔 did I miss something?

Maybe it needed some time to update. Link is working for me now: https://moon-ci-docs.huggingface.co/docs/api-inference/pr_1440/en/index

[Wauplin]

This is not expected (i.e. having import requests ... before from huggingface_hub import InferenceClient). I realized that https://huggingface.co/meta-llama/Llama-3.2-11B-Vision-Instruct?inference_api=true has a problem. Model doesn't have a chat template and therefore is not tagged as "conversational" which creates this weird side effect.

So I see 3 independent things to correct here:

1. it would be nice to recommend meta-llama/Llama-3.2-11B-Vision-Instruct first on the image-text-to-image task page (to update here)
2. we should fix in moon-landing the "conversational" detection. At the moment, it's based only on the presence of a chat template. However for idefics chatty 8b it seems it's using "use_default_system_prompt": true instead. @Rocketknight1 is it safe to assume that a model with no chat template but this parameter set to True is in fact a conversational model? And if not, which parameter could we check?
3. For non-conversational image-text-to-text (does that even exist?), we should fix the snippet generator so that only the requests-based snippet is displayed instead of this weird combination.

cc @osanseviero as well for viz'

[hanouticelina]

just to add, HuggingFaceM4/idefics2-8b-chatty has thechat_template defined in the processor_config.json. The tokenizer.chat_template attribute is supposed to be saved in tokenizer_config.json file. I guess the template was set using transformers.ProcessorMixin instead.

[hanouticelina]

for the 3rd point, pinging @mishig25 since it's related to huggingface.js/pull/938. do you think it's okay to map image-text-to-text to snippetBasic instead and define the task input here ?

[Wauplin]

Thanks @hanouticelina ! Now that huggingface/huggingface.js#956 has been merged, can you regenerate the docs which should finally highlight Llama 11B Vision model 😄

[hanouticelina]

@Wauplin I regenerated the docs and Llama 11B Vision is finally highlighted!
I also changed the sentence used in the "Recommended Models" section to ensure it works whether or not there are recommended models listed in the section:

- This is only a subset of the supported models. Find the model that suits you best [here](...)
+ Explore all available models and find the one that suits you best [here](...)

I think we're pretty much good to finally merge this PR :)

[Wauplin]

Such a nice addition! Thanks @hanouticelina for packing this up 🤗