NutriKen – Clinical Bioinformatics Platform for Latin America (Request for Infrastructure Support)

For the details, I think it would be safest to ask Hugging Face directly (website@huggingface.co), but for the infrastructure part, I think applying for a Spaces Community Grant is probably the right path :

---

A few practical suggestions, since this request seems to involve two slightly different things:

1. temporary infrastructure support for the Space , and
2. official sponsorship / endorsement / logo usage.

Those should probably be handled separately.

1. For the infrastructure request: use the Space’s Community Grant / hardware request flow

For the temporary CPU/GPU/tier upgrade, the most relevant path seems to be the Community GPU Grant / hardware upgrade request available from the Space settings.

Hugging Face’s Spaces docs mention that Spaces start on free CPU hardware by default, and that upgraded hardware can be requested from the Settings tab of the Space. The same docs also say that if a project needs help covering hardware upgrade costs, the owner can apply for a community GPU grant from the Space’s Settings tab:

• Spaces Overview — Hardware resources
• Using GPU Spaces

There is also a useful prior HF staff comment here:

• Example staff comment: use the “Apply for a community grant” button in Space Settings

In that thread, HF staff explained that using the button is important because it sends a notification to the internal review channel where staff members can see the request. So I would not rely only on a forum post for this part.

The route I would try is:

Your Space
→ Settings
→ Hardware
→ select/request the needed hardware
→ Apply for a community grant

If the UI has changed slightly, the same general idea should still apply: submit the request from the Space’s own hardware/settings page rather than only posting in the forum.

2. Be specific about what resource is actually needed

Since NutriKen is described as a Docker SDK Space , I would recommend being very explicit about the bottleneck.

From your description, the request says “CPU/GPU/tiers”, but those are different problems:

• if the issue is many clinicians opening the app at once, the bottleneck may be CPU/RAM/replicas , not GPU;
• if the issue is slow model inference, then a GPU tier may be needed;
• if the issue is cold starts during the workshop, then an upgraded always-on Space / sleep-time setting may help;
• if the issue is traffic distribution, then replicas may be relevant, but replicas are tied to upgraded hardware;
• if the issue is local cache or generated reports, then persistent storage or external storage design may matter.

Relevant docs:

• Docker Spaces
• Using GPU Spaces
• Manage your Space with huggingface_hub
• Spaces Overview — Secrets and environment variables
• Spaces Persistent Storage

A good grant request would probably include something like:

Requested support:
- Temporary hardware upgrade for the Space during the APENUT workshop
- Dates: <start date> to <end date>
- Expected total participants: <number>
- Expected concurrent users: <number>
- Current bottleneck observed: <CPU/RAM/GPU/cold-start/API latency/etc.>
- Requested tier: <CPU Upgrade / T4 / L4 / A10G / etc.>
- Why this tier: <short technical reason>
- Whether GPU is required: <yes/no + reason>
- Whether patient-identifiable data is processed or stored: <yes/no + explanation>
- Post-event plan: downgrade the Space after the 3-day workshop

That kind of detail should make the request much easier to review than a general “please upgrade the Space” request.

3. ZeroGPU may not be the right assumption here

One possible source of confusion: ZeroGPU and regular Spaces hardware grants are related but not identical paths.

ZeroGPU is a shared GPU infrastructure for Spaces. The current ZeroGPU docs describe it as being used for Gradio SDK Spaces, with dynamic GPU allocation:

• Spaces ZeroGPU docs

Since NutriKen appears to be a Docker/FastAPI-style Space, I would not assume ZeroGPU is the correct target unless Hugging Face staff says so. For a Docker Space, a normal hardware upgrade request — for example CPU Upgrade, T4, L4, A10G, etc. — may be more appropriate.

So in the grant application, I would phrase the request as:

We are requesting the most appropriate temporary Spaces hardware support for a 3-day clinical education workshop. If GPU is not necessary for this Docker/FastAPI workload, a temporary CPU/RAM upgrade would also be acceptable.

This gives the HF team room to assign the appropriate resource instead of forcing the request into the wrong category.

4. The endorsement / logo part should be treated separately

For the “official endorsement badge/logo” or “Infrastructure Sponsor” wording, I would be more cautious.

Using Hugging Face as a hosting platform is one thing. Saying that Hugging Face is an official infrastructure sponsor , or displaying an endorsement-style badge/logo, is a stronger claim. I would wait for explicit confirmation from Hugging Face before using language that implies sponsorship or endorsement.

For basic brand material, Hugging Face has a brand page:

• Hugging Face Brand Assets

But for anything like:

Official Hugging Face Infrastructure Sponsor

or:

Supported by Hugging Face

or:

Hugging Face-endorsed clinical platform

I would not use that wording unless HF explicitly approves it.

A safer wording while waiting would be something like:

Hosted on Hugging Face Spaces

or:

Deployed on Hugging Face Spaces

Those describe the infrastructure factually without implying official endorsement.

5. For a medical / clinical decision-support project, include the safety boundary clearly

Because this is a clinical bioinformatics / clinical decision-support platform, I would also include a clear safety note in the grant request and in the Space README.

Something like:

NutriKen is intended for clinical education and decision support. It does not replace professional medical judgment, individualized patient evaluation, local clinical guidelines, or review by a qualified healthcare professional. The evidence shown by the system should be verified by the clinician before any intervention.

This is especially important because the forum post uses phrases like “clinical-grade reports.” That may be understandable from the project perspective, but for infrastructure review it is safer to clearly state the boundary between:

• educational / decision-support use,
• clinical workflow support,
• and actual regulated clinical deployment.

If the Space processes real patient data, I would also clarify:

The workshop demo will not store patient-identifiable information.

or, if it does handle sensitive data:

The project handles <type of data>, stored in <location>, with <privacy/security controls>.

For a public Hugging Face Space, I would be especially careful not to process protected health information unless the privacy/security model has been reviewed.

6. Suggested short application text

You could adapt something like this for the Community Grant request:

Hello Hugging Face team,

I am requesting a temporary Spaces hardware grant for NutriKen:

https://huggingface.co/spaces/Kenryu007/Nutriken

NutriKen is an open clinical bioinformatics / nutrition decision-support platform in Spanish, built for healthcare professionals in Latin America. It integrates public evidence sources such as MSKCC, NCBI Gene, Ensembl, KEGG and PubMed, and generates structured Spanish-language educational reports for clinician review.

The project is built as a Docker SDK Space using a FastAPI backend and a lightweight web frontend:

https://github.com/abrangel/Nutriken

I have been invited by the Peruvian Academy of Nutrition (APENUT) to use NutriKen as live training infrastructure during a 3-day international workshop for hundreds of clinicians.

Requested support:
- Temporary hardware upgrade for the workshop period: <dates>
- Expected total participants: <number>
- Expected concurrent users: <number>
- Current bottleneck: <CPU/RAM/GPU/cold start/API latency/etc.>
- Requested tier: <CPU Upgrade / T4 / L4 / A10G / other>
- Duration: <number of days>
- Post-event plan: downgrade the Space after the workshop

If GPU is not the right fit for this Docker/FastAPI workload, a temporary CPU/RAM upgrade would also be very helpful.

Safety note:
NutriKen is intended for clinical education and decision support. It does not replace professional medical judgment, individualized patient evaluation, or review by a qualified healthcare professional. The evidence shown by the system should be verified by clinicians before any intervention.

Thank you for considering this request.

7. My overall recommendation

So my recommendation would be:

• keep the forum post as a public project-introduction thread;
• submit the actual infrastructure request through the Space’s Apply for a community grant flow;
• specify whether the bottleneck is CPU, RAM, GPU, cold start, or concurrent traffic;
• request a clearly time-bounded upgrade for the 3-day APENUT workshop;
• avoid implying official HF sponsorship or endorsement unless Hugging Face explicitly approves that wording;
• use factual wording like “hosted on Hugging Face Spaces” unless/until HF confirms sponsorship language.

The project seems well aligned with the spirit of community infrastructure support: open tooling, Spanish-language clinical education, Latin American access, and a time-limited workshop need. The strongest version of the request would be a concrete, reviewable hardware grant application rather than only a general forum appeal.