File size: 3,484 Bytes
714e085 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 |
# Trending Models Dashboard
A Hugging Face Space dashboard for visualizing and analyzing trending model data from the `hf-azure-internal/trending-models-analysis` dataset.
## Overview
This dashboard provides an interactive interface to explore trending machine learning models, their metadata, and their compliance status. Users can filter by date and view models organized by status or trending rank.
## Data Source
The dashboard uses data from: [`hf-azure-internal/trending-models-analysis`](https://huggingface.co/datasets/hf-azure-internal/trending-models-analysis)
### Dataset Schema
The `models` section contains the following columns:
| Column | Type | Description |
|--------|------|-------------|
| `id` | string | Model identifier |
| `trending_rank` | integer | Current trending position |
| `author` | string | Model author/organization |
| `tags` | array | Model tags |
| `license` | string | Model license type |
| `library_name` | string | Framework used (e.g., transformers, diffusers) |
| `gated` | boolean | Whether model access is gated |
| `task` | string | Model task type |
| `is_in_catalog` | boolean | Whether model is in catalog |
| `is_custom_code` | boolean | Whether model uses custom code |
| `is_excluded_org` | boolean | Whether organization is excluded |
| `is_supported_license` | boolean | Whether license is supported |
| `is_supported_library` | boolean | Whether library is supported |
| `is_safetensors` | boolean | Whether model uses safetensors |
| `is_supported_task` | boolean | Whether task is supported |
| `is_securely_scanned` | boolean | Whether model is securely scanned |
| `collected_at` | datetime | Data collection timestamp |
| `model_status` | string | Status: "to add", "blocked", or "added" |
## Features
### User Controls
- **Date Filter**: Select a specific date to filter models by `collected_at` timestamp
- **Display Mode**: Choose between two viewing modes:
- **By Status**: Organize models into three sections based on `model_status`
- **By Trending Rank**: Display all models sorted by `trending_rank` (ascending)
### Display Sections
#### By Status Mode
Models are organized into three sections:
- **To Add**: Models with `model_status = "to add"`
- **Blocked**: Models with `model_status = "blocked"`
- **Added**: Models with `model_status = "added"`
#### By Trending Rank Mode
- Single section displaying all models sorted by `trending_rank`
### Model Cards
Each model displays:
- **Model ID**: Raw value from `id` column
- **Status Indicators**: Boolean fields shown as emojis
- 🟢 Green: `true` values
- 🔴 Red: `false` values
- Fields: `is_in_catalog`, `is_custom_code`, `is_excluded_org`, `is_supported_license`, `is_supported_library`, `is_safetensors`, `is_supported_task`
- **Model Status**: Displayed in bold text
### Interactive Details
Clicking on any model card reveals additional information:
- **License Details**: Full license information
- **Task Information**: Specific task type
- **Library Details**: Framework/library used
- **Author Information**: Model author/organization
## Technical Requirements
- Will be hosted as a Hugging Face Spaces
- Interactive web interface
- Real-time data filtering
- Smooth animations for detail views
- Responsive design
## Development
This project will be implemented as a Hugging Face Space using:
- Python backend for data processing
- Interactive web framework (Gradio/Streamlit)
- Dataset integration via Hugging Face Hub |