Feature Request: Add method to get access request counts for analytics/monitoring

[balanprasanth]

Feature Description

Add API methods to retrieve access request counts (accepted/pending/rejected) without full iteration, optimized for analytics and monitoring use cases.

Context

As discussed in #3591, there's currently no way to get access counts without iterating through all users. For large gated models, this approach is inefficient for common monitoring workflows.

Motivation & Use Cases

Primary Use Cases:

1. 📊 Analytics Dashboards: Display current user counts across multiple models
2. 📈 Growth Tracking: Monitor daily/weekly user acquisition trends

Example Scenario:

# Current approach - takes 5-10 minutes 
for model in my_10_gated_models:
    count = sum(1 for _ in api.list_accepted_access_requests(model, "model"))
    dashboard.update(model, count)
# Total time: 50-100 minutes for 10 models

Proposed API

Option 1: Stats method with accuracy tradeoff

stats = api.get_access_request_stats(
    repo_id="google/gemma-2-2b-it",
    repo_type="model",
    approximate=True  # Allow slightly stale counts for performance
)
# Returns: {
#     "accepted": 595174,
#     "pending": 1234,
#     "rejected": 567,
#     "total": 596975,
#     "last_updated": "2024-12-05T10:30:00Z"  # Cache timestamp
# }

Priority

I understand this may not be short-term priority. it would be valuable for the ecosystem. Many organizations building on HF Hub would benefit from this.

Thanks for considering! 🙏

---

Related: #3591 (where this need was identified)

[julien-c]

Out of curiosity, what gated models are you interested in? (if you can share)

[balanprasanth]

FYI, google/gemma-2-9b-it models