Installation conflict

[wuhuateng]

1.Set Python to 3.8, install Unidepth, and after running bash compile. sh, running Python./scripts/demo. py will result in an error because the Python version is too low and different from Unidepth's requirement of 3.10+.
2.I noticed that there is no unidepth item in DetAny3D's requirements. txt. Did your unidepth installation succeed, or why is there no unidepth in the requirements. txt? Is unidepth installation necessary?
3.I installed it completely according to Detany3d's requirements. txt and trained it after installation, but I still encountered some issues torch.distributed.elastic.multiprocessing.errors.ChildFailedError. Is it because the GPU model is incorrect? May I ask what model of GPU you have?

[jjxjiaxue]

Hi,

Sorry for the confusion!

We are using A100 GPUs. You don’t have to install Unidepth explicitly, but some of its dependencies might still be required — so if any package is missing, you can install it via pip (e.g., pip install missing_package).

If torchrun is giving you errors, could you try running deploy.py? That might help us isolate whether the issue is related to distributed training or something else.

Happy to help further.

[wuhuateng]

Thank you sooooooooooo much! Thank you very much for your warm reply. Do I need when installing unideep,? or do I only need to install the packages that unideep depends on without compiling compile.sh?

[jjxjiaxue]

You're very welcome!

You don’t need to compile Unidepth — it’s not required for DetAny3D. You can simply install the packages that Unidepth depends on.

Alternatively, you can try running the code directly first. If there are any missing dependencies, just install them one by one via pip (e.g., pip install package_name).

[wuhuateng]

Thank you very much for your help. I still got an error during training, and the location of the error is cfg.scaler.scale(loss_total).backward()， It died directly during backway() without any error message. Due to GPU limitations, I cannot set the CUDA and Torch versions too low. Here is the pip list:
absl-py 2.3.1
addict 2.4.0
aiofiles 23.2.1
annotated-types 0.7.0
anyio 4.9.0
asttokens 3.0.0
attrs 25.3.0
blinker 1.9.0
certifi 2025.7.14
charset-normalizer 3.4.2
click 8.2.1
coloredlogs 15.0.1
comm 0.2.2
ConfigArgParse 1.7.1
contourpy 1.3.2
cycler 0.12.1
dash 3.1.1
decorator 5.2.1
defusedxml 0.7.1
einops 0.8.0
executing 2.2.0
fastapi 0.116.1
fastjsonschema 2.21.1
ffmpy 0.6.0
filelock 3.13.1
Flask 3.1.1
flatbuffers 25.2.10
fonttools 4.59.0
fsspec 2024.6.1
gradio 4.44.1
gradio_client 1.3.0
gradio_image_prompter 0.1.0
groundingdino 0.1.0 /home/wht/GroundingDINO
grpcio 1.73.1
h11 0.16.0
h5py 3.11.0
hf-xet 1.1.5
httpcore 1.0.9
httpx 0.28.1
huggingface-hub 0.33.4
humanfriendly 10.0
idna 3.10
imageio 2.34.1
importlib_metadata 8.7.0
importlib_resources 6.5.2
ipython 9.4.0
ipython_pygments_lexers 1.1.1
ipywidgets 8.1.7
itsdangerous 2.2.0
jedi 0.19.2
Jinja2 3.1.4
joblib 1.5.1
jsonschema 4.24.0
jsonschema-specifications 2025.4.1
jupyter_core 5.8.1
jupyterlab_widgets 3.0.15
kiwisolver 1.4.8
Markdown 3.8.2
markdown-it-py 3.0.0
MarkupSafe 2.1.5
matplotlib 3.7.5
matplotlib-inline 0.1.7
mdurl 0.1.2
mmcv 2.2.0
mmengine 0.10.7
mpmath 1.3.0
narwhals 1.47.0
nbformat 5.10.4
nest-asyncio 1.6.0
networkx 3.3
numpy 1.23.5
nvidia-cublas-cu12 12.1.3.1
nvidia-cuda-cupti-cu12 12.1.105
nvidia-cuda-nvrtc-cu12 12.1.105
nvidia-cuda-runtime-cu12 12.1.105
nvidia-cudnn-cu12 9.1.0.70
nvidia-cufft-cu12 11.0.2.54
nvidia-curand-cu12 10.3.2.106
nvidia-cusolver-cu12 11.4.5.107
nvidia-cusparse-cu12 12.1.0.106
nvidia-nccl-cu12 2.20.5
nvidia-nvjitlink-cu12 12.1.105
nvidia-nvtx-cu12 12.1.105
onnx 1.16.1
onnxruntime 1.18.0
open3d 0.19.0
opencv-python 4.10.0.84
orjson 3.11.0
packaging 25.0
pandas 2.0.3
parso 0.8.4
pexpect 4.9.0
pillow 10.4.0
pip 25.1
platformdirs 4.3.8
plotly 6.2.0
prompt_toolkit 3.0.51
protobuf 4.25.5
ptyprocess 0.7.0
pure_eval 0.2.3
pycocotools 2.0.10
pydantic 2.11.7
pydantic_core 2.33.2
pydub 0.25.1
Pygments 2.19.2
pyparsing 3.2.3
pyquaternion 0.9.9
python-box 7.3.2
python-dateutil 2.9.0.post0
python-multipart 0.0.20
pytz 2025.2
PyYAML 6.0.2
referencing 0.36.2
regex 2024.11.6
requests 2.32.4
retrying 1.4.0
rich 14.0.0
rpds-py 0.26.0
ruff 0.12.4
safetensors 0.5.3
scikit-learn 1.7.0
scipy 1.10.0
semantic-version 2.10.0
setuptools 78.1.1
shapely 2.1.1
shellingham 1.5.4
six 1.17.0
sniffio 1.3.1
stack-data 0.6.3
starlette 0.47.1
supervision 0.26.0
sympy 1.13.3
tabulate 0.9.0
tensorboard 2.20.0
tensorboard-data-server 0.7.2
termcolor 2.4.0
terminaltables 3.1.10
threadpoolctl 3.6.0
timm 1.0.7
tokenizers 0.21.2
tomlkit 0.12.0
torch 2.4.1+cu121
torchaudio 2.4.1+cu121
torchvision 0.19.1+cu121
tqdm 4.65.2
traitlets 5.14.3
transformers 4.53.2
trimesh 2.35.39
triton 3.0.0 1
typer 0.16.0
typing_extensions 4.12.2
typing-inspection 0.4.1
tzdata 2025.2
urllib3 2.5.0
uvicorn 0.35.0
wcwidth 0.2.13
websockets 12.0
Werkzeug 3.1.3
wheel 0.45.1
widgetsnbextension 4.0.14
xformers 0.0.28.post1
yapf 0.43.0
zipp 3.23.0
Can you see the reason for the direct death during backward()?

[jjxjiaxue]

pytorch version