openteamsinc
pypi/inspect-evals
Risk Profile
Package
Source
Risk Profile
Maturity:
Mature
Health:
Healthy
Legal:
Moderate Risk
License has additional text that may require review before use
Package was not published with a license
Security:
Healthy
Package
pypi
inspect-evals
Version
0.9.0
Last Release Date
3 days ago
License
Not specified
Dependencies
backoff
>=2.2.0
datasets
<4.7.0, >=4.5.0
huggingface_hub
>=1.2.0
hf_xet
inspect_ai
>=0.3.158
jinja2
numpy
>=1.26.0
pillow
>=11.3.0
pydantic
>=2.10.0
pyyaml
>=5.1.0
requests
>=2.32.0
tiktoken
>=0.11.0
toml
>=0.10.2
b3:
openai
rouge_score
tenacity
click
python-dotenv
agentdojo:
pydantic
[email]
deepdiff
bfcl:
mpmath
swe-bench:
swebench
>=3.0.15
docker
jsonlines
swe-lancer:
docker
types-docker
math:
sympy
antlr4-python3-runtime
~=4.11.0
worldsense:
pandas
mind2web:
beautifulsoup4
types-beautifulsoup4
lxml
lxml-stubs
sevenllm:
jieba
==0.42.1
sentence_transformers
>=5.1.1
rouge
==1.0.1
tf-keras
scicode:
gdown
h5py
scipy
sympy
ifeval:
instruction_following_eval
langdetect
ahb:
matplotlib
medqa:
bioc
niah:
pandas
kernelbench:
kernelbench
core-bench:
scipy
healthbench:
scikit-learn
personality:
huggingface-hub
sciknoweval:
nltk
rouge_score
rdkit
rdchiral
gdown
gensim
scipy
gdm-stealth:
tabulate
scipy
immutabledict
pandas
python-dateutil
cybench:
inspect-cyber
==0.1.0
cybergym:
inspect-cyber
>=0.1.0
bold:
detoxify
vaderSentiment
transformers
>=5.0.0
torch
makemesay:
nltk
abstention-bench:
scikit-learn
hydra-core
>=1.4.0.dev1
omegaconf
>=2.4.0.dev2
torch
loguru
gdown
jsonlines
gaia:
filelock
osworld:
filelock
vimgolf:
vimgolf
==0.5.1
vimgolf-challenges:
vimgolf
==0.5.1
ifevalcode:
tree-sitter
tree-sitter-cpp
gdpval:
huggingface_hub
[cli]
agentic-misalignment:
bs4
cje:
cje-eval
theagentcompany:
inspect_cyber
pandas
openpyxl
odfpy
>=1.4.1
novelty-bench:
transformers
>=4.57.1
torch
>=2.9.1
accelerate
>=1.11.0
protobuf
>=6.33.1
sentencepiece
>=0.2.1
gdm-capabilities:
google-genai
>=1.56.0
rich
python-dateutil
gdm-self-proliferation:
rich
scbench:
inspect-swe
paperbench:
drain3
test:
anthropic
openai
>=2.26.0
inspect_evals
[abstention_bench,agentdojo,b3,bold,core_bench,cybench,cybergym,fortress,gdm_capabilities,gdm_self_proliferation,gdpval,ifeval,ifevalcode,mind2web,novelty_bench,paperbench,scbench,sciknoweval,sevenllm,stealth,swe_bench,swe_lancer,vimgolf]
inspect_evals
[kernelbench]
doc:
quarto-cli
jupyter
dist:
twine
build
Source
Location
UKGovernmentBEIS/inspect_evals
Last Source Update
about 1 hour ago
Licenses
MIT License
MIT License
(text added)
MIT License
Distribution Destinations
pypi/inspect-evals
