# Solving Torch Errors

## Undefined \`\_\_nvJitLinkComplete\_12\_4\`

### Error

{% code overflow="wrap" %}

```bash
>>> import torch
  File "/home/tk/Documents/mambaforge/envs/torch/lib/python3.10/site-packages/torch/__init__.py", line 367, in <module>
    from torch._C import *  # noqa: F403
ImportError: /home/tk/Documents/mambaforge/envs/depth-pro/lib/python3.10/site-packages/torch/lib/../../nvidia/cusparse/lib/libcusparse.so.12: undefined symbol: __nvJitLinkComplete_12_4, version libnvJitLink.so.12
```

{% endcode %}

### Solution

Add the `nvjtlink` directory of the corresponding python package to link library path.

{% code overflow="wrap" %}

```bash
export LD_LIBRARY_PATH=$CONDA_PREFIX/lib/python3.10/site-packages/nvidia/nvjitlink/lib/:$LD_LIBRARY_PATH
```

{% endcode %}


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://tk233.gitbook.io/notes/ml-rl/setting-up-nvidia-tools/solving-torch-errors.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.