PyRank
stanford-crfm

nvidia-crfm-helm

Holistic Evaluation of Language Models (HELM) is an open source Python framework created by the Center for Research on Foundation Models (CRFM) at Stanford for holistic, reproducible and transparent evaluation of foundation models, including large language models (LLMs) and multimodal models.
PyPI GitHub
More links
Download Stats
Monthly downloads
102
Weekly downloads
43
GitHub stars *
2.9K

Downloads in the 30 and 7 days to October 10, 2026. Refreshed weekly.

Package Information

Latest Version
25.10
Requires Python
>=3.9
Licence
Apache License 2.0
Last Release
11 months ago
Author
Stanford CRFM, NVIDIA
First Release
August 6, 2025
Total Releases
6
Python Versions
3 3.9 3.10 3.11
No reported malware or high/critical advisories · Details

The current version has no OSV.dev malware reports and no critical or high-severity vulnerabilities.

Package Insights

What's this?
Score breakdown - each component scores 0-100, weighted by percentage
Maintenance
45
Popularity
74
Adoption
20
Community
80
Provenance
0
How the maintenance status was determined
Last commit
133 days ago
Monthly downloads
102
Active
Activity within 30 days
• Stable
Activity within 31-180 days
Minimal
Activity within 181-365 days, or no activity data
At Risk
No activity for > 365 days, but > 10,000 monthly downloads
Archived
Repository flagged as archived on GitHub
How the popularity category was determined
Monthly downloads
102
Stars
2,939
Stars per 1k downloads
28813.73
Industry Standard
> 10,000,000 monthly downloads
Hidden Gem
< 0.5 stars per 1k downloads and > 100,000 monthly downloads
Community Favorite
> 5 stars per 1k downloads
Rising Star
Balanced ratio of stars to downloads
• Unknown
Fewer than 1,000 monthly downloads, or no star data
Needs at least 1,000 monthly downloads to classify popularity.
How download momentum is calculated
Week daily avg
6
Month daily avg
3
Momentum
—
((week_daily_avg - month_daily_avg) / month_daily_avg) * 100
Momentum not reported for packages with < 5,000 monthly downloads — percentages are noisy at low volume.
More insights and metrics
Python version support tiers
requires_python
>=3.9
Cutting Edge
Minimum supported version is 3.14 or newer
Modern
Minimum supported version is 3.11 to 3.13
• Legacy
Supports 3.10 or older (or Python 2)
Unknown
No requires_python specified or unparseable
License categories
License
Apache License 2.0
• Permissive
MIT, Apache, BSD, ISC, Unlicense, CC0, etc.
Copyleft
GPL, LGPL, AGPL, MPL, EPL, etc.
Proprietary
Commercial, all rights reserved, custom EULAs
Unknown
No license specified, or none of the above match
How the release cadence was determined
Last release
350 days ago
Total releases
6
Releases/year
5.0
Rapid
Last release within 365 days and > 12 releases/year
• Regular
Last release within 365 days
Slow
Last release 366-730 days ago
Stalled
Last release > 730 days ago, or no releases
Unknown
No usable release date available
How contributor risk was determined
Contributors
181
Monthly downloads
102
High Load
≤ 2 contributors and ≥ 50,000 monthly downloads
Moderate Load
< 10 contributors and ≥ 500,000 monthly downloads
Low Load
< 25 contributors and ≥ 5,000,000 monthly downloads, or sole maintainer with ≥ 10,000 monthly downloads
• Balanced
Enough contributors for the package's popularity
Unknown
No contributor data available
Engagement
69
 
Issue Burden
23.1
per 1k stars
Discussions
0.0
per 1k stars
Fork Ratio
42.8%
contrib / forks
Stale Release
217d
 
PR Velocity
Quiet
 
Not all downloads include detailed metadata. Breakdowns reflect the subset that does.

Downloads by OS

No OS data available

Python Versions

No Python version data available

Top Countries

1. FR FR
3
2. SE SE
3

Weekly Downloads

GitHub Stars *

Dependencies

32
113 optional dependencies
  • accelerate~=0.25[accelerate]
    extra == "accelerate"
  • ai2-olmo~=0.2[allenai]
    extra == "allenai"
  • aiohttp<4.0.0,>=3.12.13[medhelm]
    extra == "medhelm"
  • aleph-alpha-client~=2.14.0[aleph-alpha]
    extra == "aleph-alpha"
  • anthropic~=0.48[anthropic]
    extra == "anthropic"
  • anthropic<1.0.0,>=0.57.1[medhelm]
    extra == "medhelm"
  • awscli~=1.33.13[amazon]
    extra == "amazon"
  • bert-score~=0.3[summarization]
    extra == "summarization"
  • bert-score<1.0.0,>=0.3.13[medhelm]
    extra == "medhelm"
  • boto3~=1.34.131[amazon]
    extra == "amazon"
  • botocore~=1.34.131[amazon]
    extra == "amazon"
  • bottle<1.0.0,>=0.12.25[medhelm]
    extra == "medhelm"
  • cohere~=5.3[cohere]
    extra == "cohere"
  • colorcet~=3.0.1[plots]
    extra == "plots"
  • datasets<3.0.0,>=2.21.0[medhelm]
    extra == "medhelm"
  • einops~=0.7.0[vlm]
    extra == "vlm"
  • einops-exts~=0.0.4[vlm]
    extra == "vlm"
  • evaluate~=0.4.1[unitxt]
    extra == "unitxt"
  • fairlearn~=0.9.0[decodingtrust]
    extra == "decodingtrust"
  • flask<4.0.0,>=3.0.3[medhelm]
    extra == "medhelm"
  • gdown~=5.1[scenarios]
    extra == "scenarios"
  • gin-config<1.0.0,>=0.5.0[medhelm]
    extra == "medhelm"
  • google-api-python-client~=2.64[metrics]
    extra == "metrics"
  • google-cloud-aiplatform~=1.48[google]
    extra == "google"
  • gradio-client~=1.3[metrics]
    extra == "metrics"
  • gunicorn>=20.1[proxy-server]
    extra == "proxy-server"
  • httpx<0.28.0[anthropic]
    extra == "anthropic"
  • huggingface-hub<1.0.0,>=0.33.1[medhelm]
    extra == "medhelm"
  • immutabledict~=4.2.0[metrics]
    extra == "metrics"
  • jieba~=0.42.1[cleva]
    extra == "cleva"
  • langdetect~=1.0.9[metrics]
    extra == "metrics"
  • langdetect~=1.0[cleva]
    extra == "cleva"
  • lxml<7.0.0,>=6.0.0[medhelm]
    extra == "medhelm"
  • matplotlib~=3.6.0[plots]
    extra == "plots"
  • matplotlib<4.0.0,>=3.10.3[medhelm]
    extra == "medhelm"
  • mistralai~=1.1[mistral]
    extra == "mistral"
  • nltk<4.0.0,>=3.9.1[medhelm]
    extra == "medhelm"
  • numba~=0.56[metrics]
    extra == "metrics"
  • numpy<2.0.0,>=1.26.4[medhelm]
    extra == "medhelm"
  • open-clip-torch~=2.24[vlm]
    extra == "vlm"
  • openai~=1.70[openai]
    extra == "openai"
  • openai<2.0.0,>=1.93.0[medhelm]
    extra == "medhelm"
  • opencc~=1.1[cleva]
    extra == "cleva"
  • openpyxl<4.0.0,>=3.1.5[medhelm]
    extra == "medhelm"
  • pandas<3.0.0,>=2.3.0[medhelm]
    extra == "medhelm"
  • parameterized<1.0.0,>=0.9.0[medhelm]
    extra == "medhelm"
  • pillow<12.0.0,>=11.3.0[medhelm]
    extra == "medhelm"
  • pillow~=10.2[images]
    extra == "images"
  • protobuf<7.0.0,>=6.31.1[medhelm]
    extra == "medhelm"
  • psutil<7.0.0,>=6.0.0[medhelm]
    extra == "medhelm"
  • pyarrow<21.0.0,>=20.0.0[medhelm]
    extra == "medhelm"
  • pyarrow-hotfix<1.0.0,>=0.7[medhelm]
    extra == "medhelm"
  • pydantic~=2.0[openai]
    extra == "openai"
  • pydantic<3.0.0,>=2.11.7[medhelm]
    extra == "medhelm"
  • pymongo~=4.2[mongo]
    extra == "mongo"
  • pyonmttok==1.37.0[seahelm]
    extra == "seahelm"
  • pypinyin~=0.49.0[cleva]
    extra == "cleva"
  • pythainlp==5.0.0[seahelm]
    extra == "seahelm"
  • python-crfsuite~=0.9.11[seahelm]
    extra == "seahelm"
  • python-dateutil<3.0.0,>=2.9.0.post0[medhelm]
    extra == "medhelm"
  • python-docx<2.0.0,>=1.2.0[medhelm]
    extra == "medhelm"
  • python-dotenv<2.0.0,>=1.1.1[medhelm]
    extra == "medhelm"
  • pytrec-eval==0.5[ranking]
    extra == "ranking"
  • pyyaml<7.0.0,>=6.0.2[medhelm]
    extra == "medhelm"
  • regex<2025.0.0,>=2024.11.6[medhelm]
    extra == "medhelm"
  • reka-api~=2.0.0[reka]
    extra == "reka"
  • requests<3.0.0,>=2.32.4[medhelm]
    extra == "medhelm"
  • retrying<2.0.0,>=1.4.0[medhelm]
    extra == "medhelm"
  • rich<15.0.0,>=14.0.0[medhelm]
    extra == "medhelm"
  • rouge-score<1.0.0,>=0.1.2[medhelm]
    extra == "medhelm"
  • sacrebleu~=2.2.1[seahelm]
    extra == "seahelm"
  • sacrebleu~=2.2.1[metrics]
    extra == "metrics"
  • scaleapi~=2.13.0[human-evaluation]
    extra == "human-evaluation"
  • scikit-learn<2.0.0,>=1.7.0[medhelm]
    extra == "medhelm"
  • scipy~=1.10[vlm]
    extra == "vlm"
  • scipy<2.0.0,>=1.10.0[medhelm]
    extra == "medhelm"
  • seaborn~=0.11.0[plots]
    extra == "plots"
  • sentencepiece<1.0.0,>=0.2.0[medhelm]
    extra == "medhelm"
  • sentencepiece~=0.2.0[yandex]
    extra == "yandex"
  • setuptools<81.0.0,>=80.9.0[medhelm]
    extra == "medhelm"
  • simple-slurm~=0.2.6[slurm]
    extra == "slurm"
  • six<2.0.0,>=1.17.0[medhelm]
    extra == "medhelm"
  • smart-open<8.0.0,>=7.1.0[medhelm]
    extra == "medhelm"
  • spacy<4.0.0,>=3.8.7[medhelm]
    extra == "medhelm"
  • structlog<25.0.0,>=24.1.0[medhelm]
    extra == "medhelm"
  • surge-api~=1.1.0[human-evaluation]
    extra == "human-evaluation"
  • sympy~=1.11.1[scenarios]
    extra == "scenarios"
  • sympy<2.0.0,>=1.14.0[medhelm]
    extra == "medhelm"
  • tiktoken~=0.7[openai]
    extra == "openai"
  • tiktoken<1.0.0,>=0.9.0[medhelm]
    extra == "medhelm"
  • together~=1.1[together]
    extra == "together"
  • tokenizers<1.0.0,>=0.21.2[medhelm]
    extra == "medhelm"
  • tokenizers>=0.13.3[aleph-alpha]
    extra == "aleph-alpha"
  • toml<1.0.0,>=0.10.2[medhelm]
    extra == "medhelm"
  • torch<3.0.0,>=2.7.1[medhelm]
    extra == "medhelm"
  • torch~=2.1[vlm]
    extra == "vlm"
  • torchvision<3.0.0,>=0.14.1[vlm]
    extra == "vlm"
  • torchvision<1.0.0,>=0.22.1[medhelm]
    extra == "medhelm"
  • tqdm<5.0.0,>=4.67.1[medhelm]
    extra == "medhelm"
  • transformers<5.0.0,>=4.53.0[medhelm]
    extra == "medhelm"
  • transformers-stream-generator~=0.0.4[vlm]
    extra == "vlm"
  • triton<4.0.0,>=3.3.1[medhelm]
    extra == "medhelm"
  • typer<1.0.0,>=0.16.0[medhelm]
    extra == "medhelm"
  • typing-extensions<5.0.0,>=4.14.0[medhelm]
    extra == "medhelm"
  • uncertainty-calibration<1.0.0,>=0.1.4[medhelm]
    extra == "medhelm"
  • unidecode~=1.3[cleva]
    extra == "cleva"
  • websocket-client<2.0.0,>=1.3.3[medhelm]
    extra == "medhelm"
  • websocket-client~=1.3.2[anthropic]
    extra == "anthropic"
  • writerai~=4.0[writer]
    extra == "writer"
  • xlrd~=2.0.1[scenarios]
    extra == "scenarios"
  • xmltodict<1.0.0,>=0.14.2[medhelm]
    extra == "medhelm"
  • yq<4.0.0,>=3.1.1[medhelm]
    extra == "medhelm"
  • zstandard<1.0.0,>=0.18.0[medhelm]
    extra == "medhelm"

Used By

0
Not used by any tracked packages

Repository details

Forks
423
Issues
68
Last Commit
4 months ago
Contributors
181
Open PRs
46
Archived
No

Data and sources

Tracking Since
September 26, 2026
Data Updated
October 11, 2026
Data Freshness
Fresh
Keywords
language models benchmarking nvidia fork stanford crfm helm
Sources and methodology