https://github.com/lbann/llama
Parallel implementation of the LLaMA models
https://github.com/lbann/llama
Last synced: 3 months ago
JSON representation
Parallel implementation of the LLaMA models
- Host: GitHub
- URL: https://github.com/lbann/llama
- Owner: LBANN
- License: other
- Created: 2024-11-09T14:23:56.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2025-05-12T23:32:52.000Z (about 1 year ago)
- Last Synced: 2025-07-07T18:53:28.569Z (11 months ago)
- Language: Python
- Homepage:
- Size: 124 KB
- Stars: 0
- Watchers: 1
- Forks: 2
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
- Contributing: CONTRIBUTING.md
- License: LICENSE
Awesome Lists containing this project
README
# LLaMA Repository
Distributed implementation of the LLaMA 3.x model. Optimzied to allow both
pipeline and tensor parallel inference execution using PyTorch.
```
torchrun-hpc -N1 -n2 --rdv tcp chat_server.py --model-dir
```
# LBANN: Livermore Big Artificial Neural Network Toolkit
The Livermore Big Artificial Neural Network toolkit (LBANN) is an
open-source, HPC-centric, deep learning training framework that is
optimized to compose multiple levels of parallelism.
LBANN provides model-parallel acceleration through domain
decomposition to optimize for strong scaling of network training. It
also allows for composition of model-parallelism with both data
parallelism and ensemble training methods for training large neural
networks with massive amounts of data. LBANN is able to advantage of
tightly-coupled accelerators, low-latency high-bandwidth networking,
and high-bandwidth parallel file systems.
## Publications
A list of publications, presentations and posters are shown
[here](https://lbann.readthedocs.io/en/latest/publications.html).
## Reporting issues
Issues, questions, and bugs can be raised on the [Github issue
tracker](https://github.com/LBANN/lbann/issues).