can't run test after installation on windows

[Yanting-Yao]

hi. I am trying to install on Win10.
in anaconda terminal I ran:

conda create -n dlc-live python=3.7 tensorflow==1.13.1
conda activate dlc-live
pip install deeplabcut-live

all went smoothly.

When I ran dlc-live-test I got the following error. Please help. Thank you.

(dlc-live) PS C:\WINDOWS\system32> dlc-live-test
C:\Users\yyao\anaconda3\envs\dlc-live\lib\site-packages\tensorflow\python\framework\dtypes.py:526: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint8 = np.dtype([("qint8", np.int8, 1)])
C:\Users\yyao\anaconda3\envs\dlc-live\lib\site-packages\tensorflow\python\framework\dtypes.py:527: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_quint8 = np.dtype([("quint8", np.uint8, 1)])
C:\Users\yyao\anaconda3\envs\dlc-live\lib\site-packages\tensorflow\python\framework\dtypes.py:528: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint16 = np.dtype([("qint16", np.int16, 1)])
C:\Users\yyao\anaconda3\envs\dlc-live\lib\site-packages\tensorflow\python\framework\dtypes.py:529: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_quint16 = np.dtype([("quint16", np.uint16, 1)])
C:\Users\yyao\anaconda3\envs\dlc-live\lib\site-packages\tensorflow\python\framework\dtypes.py:530: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint32 = np.dtype([("qint32", np.int32, 1)])
C:\Users\yyao\anaconda3\envs\dlc-live\lib\site-packages\tensorflow\python\framework\dtypes.py:535: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
np_resource = np.dtype([("resource", np.ubyte, 1)])

Creating temporary directory...

Downloading Video to C:\Users\yyao\dlc-live-tmp\dog_clip.avi
Model already downloaded, using cached version

Running inference...

Run 1 / 1

Traceback (most recent call last):
File "C:\Users\yyao\anaconda3\envs\dlc-live\lib\runpy.py", line 193, in _run_module_as_main
"main", mod_spec)
File "C:\Users\yyao\anaconda3\envs\dlc-live\lib\runpy.py", line 85, in run_code
exec(code, run_globals)
File "C:\Users\yyao\anaconda3\envs\dlc-live\Scripts\dlc-live-test.exe_main.py", line 7, in
File "C:\Users\yyao\anaconda3\envs\dlc-live\lib\site-packages\dlclive\check_install\check_install.py", line 72, in main
benchmark_videos(str(model_dir), video_file, display=display, resize=0.5, pcutoff=0.25)
File "C:\Users\yyao\anaconda3\envs\dlc-live\lib\site-packages\dlclive\benchmark.py", line 644, in benchmark_videos
output=output,
File "C:\Users\yyao\anaconda3\envs\dlc-live\lib\site-packages\dlclive\benchmark.py", line 289, in benchmark
inf_times = np.zeros(n_frames)
ValueError: negative dimensions are not allowed

my pip list
Package Version

---

absl-py 1.3.0
astor 0.8.1
astunparse 1.6.3
attrs 24.2.0
backports.cached-property 1.0.2
build 0.10.0
CacheControl 0.12.14
cachetools 5.5.2
certifi 2022.12.7
charset-normalizer 3.4.3
cleo 2.1.0
colorama 0.4.6
colorcet 3.1.0
crashtest 0.4.1
deeplabcut-live 1.0.4
distlib 0.4.0
dlclibrary 0.0.11
dulwich 0.21.7
filelock 3.12.2
flatbuffers 25.9.23
flit_core 3.6.0
fsspec 2023.1.0
gast 0.4.0
google-auth 2.41.1
google-auth-oauthlib 0.4.6
google-pasta 0.2.0
grpcio 1.42.0
h5py 3.7.0
html5lib 1.1
huggingface-hub 0.16.4
idna 3.10
importlib-metadata 6.7.0
importlib-resources 5.12.0
installer 0.7.0
jaraco.classes 3.2.3
jsonschema 4.17.3
keras 2.11.0
Keras-Applications 1.0.8
Keras-Preprocessing 1.1.2
keyring 23.13.1
libclang 18.1.1
lockfile 0.12.2
Markdown 3.4.1
MarkupSafe 2.1.1
mkl-fft 1.3.0
mkl-random 1.1.1
mkl-service 2.3.0
mock 4.0.3
more-itertools 9.1.0
msgpack 1.0.5
numexpr 2.8.6
numpy 1.21.6
oauthlib 3.2.2
opencv-python-headless 4.12.0.88
opt-einsum 3.3.0
packaging 24.0
pandas 1.3.5
pexpect 4.9.0
Pillow 9.5.0
pip 22.3.1
pkginfo 1.10.0
pkgutil_resolve_name 1.3.10
platformdirs 3.11.0
poetry 1.5.1
poetry-core 1.6.1
poetry-plugin-export 1.4.0
protobuf 3.19.6
ptyprocess 0.7.0
py-cpuinfo 9.0.0
pyasn1 0.5.1
pyasn1-modules 0.3.0
pyproject_hooks 1.2.0
pyrsistent 0.19.3
python-dateutil 2.9.0.post0
pytz 2025.2
pywin32-ctypes 0.2.3
PyYAML 6.0.1
rapidfuzz 3.4.0
requests 2.31.0
requests-oauthlib 2.0.0
requests-toolbelt 1.0.0
rsa 4.9.1
ruamel.yaml 0.17.40
ruamel.yaml.clib 0.2.8
scipy 1.6.2
setuptools 65.6.3
shellingham 1.5.4
six 1.16.0
tables 3.7.0
tensorboard 1.13.1
tensorboard-data-server 0.6.1
tensorboard-plugin-wit 1.8.1
tensorflow 1.13.1
tensorflow-estimator 1.13.0
tensorflow-intel 2.11.0
tensorflow-io-gcs-filesystem 0.31.0
termcolor 2.1.0
tomli 2.0.1
tomlkit 0.12.5
tqdm 4.67.1
trove-classifiers 2025.9.11.17
typing_extensions 4.7.1
urllib3 1.26.20
virtualenv 20.26.6
webencodings 0.5.1
Werkzeug 2.2.2
wheel 0.38.4
wincertstore 0.2
wrapt 1.16.0
zipp 3.11.0

[Yanting-Yao]

Will anyone please share a environment.yml that works. Many thanks.

[sneakers-the-rat]

Hmm so the n_frames value that is apparently negative is computed here:

DeepLabCut-live/dlclive/benchmark.py

Lines 236 to 240 in b60f21b

	cap = cv2.VideoCapture(video_path)
	ret, frame = cap.read()
	n_frames = (
	n_frames
	if (n_frames > 0) and (n_frames < cap.get(cv2.CAP_PROP_FRAME_COUNT) - 1)

so looks like there must be some problem either with opencv or with the video file. Opencv capture properties are sort of notoriously fragile and unneccessarily platform/backend dependent, so that could be a thing. Could you check if the dog video (at location above) is correctly downloaded and plays?

[Yanting-Yao]

The video file is 183kb and can't be played with the players I've got. I've tried to delete it yet the newly downloaded video remained the same.

[sneakers-the-rat]

That would do it. The video is 1.25MB, no idea why the download might be failing, but try replacing it manually: https://github.com/DeepLabCut/DeepLabCut-live/blob/main/check_install/dog_clip.avi

[Yanting-Yao]

That would do it. The video is 1.25MB, no idea why the download might be failing, but try replacing it manually: https://github.com/DeepLabCut/DeepLabCut-live/blob/main/check_install/dog_clip.avi

I've tried that. But when I run dlc-live-test it downloads again and replaces the good video file with the 183kb broken file.

[sneakers-the-rat]

weird... it shouldn't:

DeepLabCut-live/dlclive/check_install/check_install.py

Lines 54 to 59 in b60f21b

	if not os.path.exists(video_file):
	print(f"Downloading Video to {video_file}")
	url_link = "https://github.com/DeepLabCut/DeepLabCut-live/blob/main/check_install/dog_clip.avi?raw=True"
	urllib.request.urlretrieve(url_link, video_file, reporthook=urllib_pbar)
	else:
	print(f"Video already exists at {video_file}")

ah it's probably because the package hasn't been re-released on pypi since #126