Launching local LLM llama.cpp + analysis of attention and activations (Windows)

AI & Machine Learning, Python — incorrectly specified categories?

22 USD

Project translated automatically. Log in or register, to view the original

It is necessary to install a ready-made LLM model and the software for it.

In order to:

✅ run a local LLM (without the internet)

✅ see attention heatmaps

✅ view logits

✅ see layer activations

✅ control generation parameters

✅ analyze the internal workings of the model

The software that needs to be installed for all this to work:
- CMake
- Git

- Visual Studio Build Tools
- Python

There is an installation guide available, assistance is needed with the implementation of this task. Please suggest a price

Proposals 4

Oleksandr Kovalchuk

16 0

Budget: 1000 UAH Deadline: 1 day

I can help you with the installation of this software. I often use Python, Git, and VS in my work, so I can help solve installation issues. CMake is not as common, but installing it is also not a problem. I suggest running the LLM through the Ollama utility - it can be quite conveniently tailored to your needs and provides access to many models for free.

Please let me know if you are only interested in the installation or if you also need to implement some additional program on top of the LLM? If an additional program is needed, please write in more detail about its purpose and desired functionality in private. Thank you!