bug: CUDA build workflow not installing correct CUDA version

# Prerequisites

Please answer the following questions for yourself before submitting an issue.

- [x] I am running the latest code. Development is very rapid so there are no tagged versions as of now.
- [x] I carefully followed the [README.md](https://github.com/abetlen/llama-cpp-python/blob/main/README.md).
- [x] I [searched using keywords relevant to my issue](https://docs.github.com/en/issues/tracking-your-work-with-issues/filtering-and-searching-issues-and-pull-requests) to make sure that I am creating a new issue that is not already open (or closed).
- [x] I reviewed the [Discussions](https://github.com/abetlen/llama-cpp-python/discussions), and have a new bug or useful enhancement to share.

# Expected Behavior

CUDA build workflow (`build-wheels-cuda.yaml` ) should install the requested version of CUDA toolkit required for the CUDA build process.

# Current Behaviour

Right now CUDA build workflow does not install the requested cuda-toolkit version during CUDA build process.

See below where I replicated the issue,

https://github.com/ChamalGomesHSO/llama-cpp-python/actions/runs/19285030787/job/55143931103

In short I requested 12.4.1 but Mamba installed cuda-toolkit 13.0.1 instead:

# Environment and Context

GitHub workflow (CUDA build):

To reproduce the error as shown above I restricted the build scope of the CUDA build workflow to the following (but it applied to all combinations in the matrix)

```
Python version 3.13
OS: Ubuntu 22.04
CUDA: 12.4.1
```
# Steps to Reproduce

Please provide detailed steps for reproducing the issue. We are not sitting in front of your screen, so the more detail the better.

1. Run the existing build CUDA workflow (`build-wheels-cuda.yaml` )
2. Check Mamba installation step logs
3. You will notice requested CUDA does not match with mamba installation