huggingface/community-evals

آخر نشاط مصدر مسجل: ٣ فبراير ٢٠٢٦ في ١٣:٤٢

آخر تحديث لفهرس SkillsMP: ٩ نوفمبر ٢٠٢٦ في ١٤:٢٢

skills مجمعة: ١

نجوم GitHub: ١٤
تفرعات GitHub: ٣

التثبيت بمساعدة مساعد ذكي

انسخ هذا الطلب إلى المساعد الذكي الذي تستخدمه.

Follow https://skillsmp.com/skill-install/prompt.md to install Agent Skills from https://github.com/huggingface/community-evals.

ملف README للمستودع

مصدر ملف README للمستودع: README.md

قراءة الملف الأصلي

عرض ملف README كاملًا

# community-evals A repository for community-driven evaluation results for open source models on HuggingFace Hub. This project adds structured evaluation results to model repositories using the new [`.eval_results/` format](https://huggingface.co/docs/hub/eval-results), enabling: - Benchmark Leaderboards: Results appear on both model pages and benchmark dataset leaderboards - Community Contributions: Anyone can submit evaluation results via Pull Request - Verified Results: Peer reviewed and verified evaluation results ![Model Evaluation Results](https://huggingface.co/huggingface/documentation-images/resolve/main/evaluation-results/eval-results-previw.png) ## Supported Benchmarks | Benchmark | Hub Dataset ID | |-----------|---------------| | HLE | [cais/hle](https://huggingface.co/datasets/cais/hle) | | GPQA | [Idavidrein/gpqa](https://huggingface.co/datasets/Idavidrein/gpqa) | | MMLU-Pro | [TIGER-Lab/MMLU-Pro](https://huggingface.co/datasets/TIGER-Lab/MMLU-Pro) | | GSM8K | [openai/gsm8k](https://huggingface.co/datasets/openai/gsm8k) | ## Quick Start ### Prerequisites ```bash # Set environment variables (or add to .env file) export HF_TOKEN="your-huggingface-token" # Required: write access for PRs ``` ### Add a Single Evaluation ```bash # Preview (dry run) uv run scripts/evaluation_manager.py add-eval \ --benchmark HLE \ --repo-id "moonshotai/Kimi-K2-Thinking" # Create PR uv run scripts/evaluation_manager.py add-eval \ --benchmark HLE \ --repo-id "moonshotai/Kimi-K2-Thinking" \ --create-pr ``` ### Batch Process Trending Models ```bash # Preview top 10 trending LLMs uv run scripts/batch_eval_prs.py --limit 10 --benchmark HLE --dry-run # Create PRs for models with scores uv run scripts/batch_eval_prs.py --limit 10 --benchmark HLE ``` --- ## Manual Usage (specific model and benchmark) ### Step 1: Look Up Score (Preview) ```bash # From model card (default) uv run scripts/evaluation_manager.py add-eval \ --benchmark HLE \ --repo-id "model/name" # From Artificial Analysis uv run scripts/evaluation_manager.py add-eval \ --benchmark GPQA \ --repo-id "model/name" \ --source aa ``` **Sources:** - `model_card` (default): Extract from README tables - `aa`: Query Artificial Analysis API (requires `AA_API_KEY`) - `papers`: HuggingFace Papers (not yet implemented) ### Step 2: Check Existing PRs Always check for existing PRs before creating new ones: ```bash uv run scripts/evaluation_manager.py get-prs --repo-id "model/name" ``` ### Step 3: Apply Changes ```bash # Create PR (someone else's model) uv run scripts/evaluation_manager.py add-eval \ --benchmark HLE \ --repo-id "other-user/their-model" \ --create-pr # Push directly (model authors) uv run scripts/evaluation_manager.py add-eval \ --benchmark HLE \ --repo-id "your-username/your-model" \ --apply ``` ## Manual Usage (get all evaluations for a model card) For models with evaluation tables in their README: ```bash # 1. Inspect available tables uv run scripts/evaluation_manager.py inspect-tables --repo-id "model/name" ``` This will output the available tables and their numbers which you can then use to extract the evaluations. --- ## Agentic Usage This repository includes a Claude Code skill at `.claude/skills/hugging-face-evaluation/` for automated evaluation management. ### Ideal Prompts **Add specific benchmark to a model:** ``` Add the HLE evaluation results to moonshotai/Kimi-K2-Thinking from the model card ``` **Batch process trending models:** ``` Find the top 20 trending text-generation models and add HLE scores from Model Cards. Do a dry run first. ``` ``` Find the top 20 trending text-generation models and add HLE scores from Artificial Analysis. Do a dry run first. ``` **Check and add multiple benchmarks:** ``` For meta-llama/Llama-3.1-8B-Instruct, check the model card for GPQA and MMLU-Pro scores and create PRs if found. ``` **Extract from README tables:** ``` Inspect the evaluation tables in deepseek-ai/DeepSeek-V3 and extract all benchmark scores to .eval_results/ format ``` ### Workflow for Agents 1. **Check for existing PRs** before creating new ones 2. **Preview first** (dry run) to verify scores are found 3. **Use appropriate source** (model_card for README tables, aa for Artificial Analysis) 4. **Create PRs** for models you don't own, push directly for your own ### Example Agent Session ``` User: Add HLE score to MiniMaxAI/MiniMax-M2.1 from Artificial Analysis Agent: 1. Checks for existing open PRs on MiniMaxAI/MiniMax-M2.1 2. Looks up HLE score from Artificial Analysis API 3. Finds: HLE = 22.2% 4. Creates PR with .eval_results/hle.yaml containing the score 5. Returns PR URL to user ``` --- ## Output Format Results are stored as YAML files in `.eval_results/`: ```yaml # .eval_results/hle.yaml - dataset: id: cais/hle # Hub Benchmark dataset ID task_id: default # Optional: specific task/leaderboard value: 22.2 # Metric value date: "2026-01-14" # ISO-8601 date source: # Attribution url: https://artificialanalysis.ai name: Artificial Analysis ``` See [HuggingFace Eval Results Documentation](https://huggingface.co/docs/hub/eval-results) for full format specification. --- ## Run Results Results from batch runs are saved to `runs/` directory: ``` runs/ ├── hle_20260114_abc123.json ├── gpqa_20260114_def456.json └── ... ``` Each file contains: ```json { "benchmark": "HLE", "source": "aa", "source_url": "https://artificialanalysis.ai", "created": "2026-01-14T08:00:00Z", "results": [ { "repo_id": "MiniMaxAI/MiniMax-M2.1", "value": 22.2, "status": "pr_created", "source_url": "https://artificialanalysis.ai" } ] } ``` **Status values:** `pr_created`, `uploaded`, `not_found`, `dry_run`, `error`

جارٍ تحميل ملف README للمستودع…

Skills في هذا المستودع

علماء الأحياء، جميع الآخرون

١ فئات مهنية · 100% مصنفة

skill

المهنة

الوصف

آخر تحديث

hugging-face-evaluation

علماء الأحياء، جميع الآخرون

Add evaluation results to Hugging Face model repositories using the .eval_results/ format. Uses HF CLI for PR management and manual YAML creation.

لغة النص الأصلي: الإنجليزية

2026-02-03

عرض ١ من أصل ١ skills مجمعة.