LlamaGPT

A self-hosted, offline, ChatGPT-like chatbot, powered by Llama 2. 100% private, with no data leaving your device.
New: Support for Code Llama models and Nvidia GPUs.

umbrel.com (we're hiring) »

Contents

1. Demo
2. Supported Models
3. How to install
   • On umbrelOS home server
   • On M1/M2 Mac
   • Anywhere else with Docker
   • Kubernetes
4. OpenAI-compatible API
5. Benchmarks
6. Roadmap and contributing
7. Acknowledgements

Demo

https://github.com/getumbrel/llama-gpt/assets/10330103/5d1a76b8-ed03-4a51-90bd-12ebfaf1e6cd

Supported models

Currently, LlamaGPT supports the following models. Support for running custom models is on the roadmap.

Model name	Model size	Model download size	Memory required
Nous Hermes Llama 2 7B Chat (GGML q4_0)	7B	3.79GB	6.29GB
Nous Hermes Llama 2 13B Chat (GGML q4_0)	13B	7.32GB	9.82GB
Nous Hermes Llama 2 70B Chat (GGML q4_0)	70B	38.87GB	41.37GB
Code Llama 7B Chat (GGUF Q4_K_M)	7B	4.24GB	6.74GB
Code Llama 13B Chat (GGUF Q4_K_M)	13B	8.06GB	10.56GB
Phind Code Llama 34B Chat (GGUF Q4_K_M)	34B	20.22GB	22.72GB

How to install

Install LlamaGPT on your umbrelOS home server

Running LlamaGPT on an umbrelOS home server is one click. Simply install it from the Umbrel App Store.

Install LlamaGPT on M1/M2 Mac

Make sure your have Docker and Xcode installed.

Then, clone this repo and cd into it:

git clone https://github.com/getumbrel/llama-gpt.git
cd llama-gpt

Run LlamaGPT with the following command:

./run-mac.sh --model 7b

You can access LlamaGPT at http://localhost:3000.

To run 13B or 70B chat models, replace 7b with 13b or 70b respectively. To run 7B, 13B or 34B Code Llama models, replace 7b with code-7b, code-13b or code-34b respectively.

To stop LlamaGPT, do Ctrl + C in Terminal.

Install LlamaGPT anywhere else with Docker

You can run LlamaGPT on any x86 or arm64 system. Make sure you have Docker installed.

Then, clone this repo and cd into it:

git clone https://github.com/getumbrel/llama-gpt.git
cd llama-gpt

Run LlamaGPT with the following command:

./run.sh --model 7b

Or if you have an Nvidia GPU, you can run LlamaGPT with CUDA support using the --with-cuda flag, like:

./run.sh --model 7b --with-cuda

You can access LlamaGPT at http://localhost:3000.

To run 13B or 70B chat models, replace 7b with 13b or 70b respectively. To run Code Llama 7B, 13B or 34B models, replace 7b with code-7b, code-13b or code-34b respectively.

To stop LlamaGPT, do Ctrl + C in Terminal.

Note: On the first run, it may take a while for the model to be downloaded to the /models directory. You may also see lots of output like this for a few minutes, which is normal:

llama-gpt-llama-gpt-ui-1       | [INFO  wait] Host [llama-gpt-api-13b:8000] not yet available...

After the model has been automatically downloaded and loaded, and the API server is running, you'll see an output like:

llama-gpt-ui_1   | ready - started server on 0.0.0.0:3000, url: http://localhost:3000

You can then access LlamaGPT at http://localhost:3000.

---

Install LlamaGPT with Kubernetes

First, make sure you have a running Kubernetes cluster and kubectl is configured to interact with it.

Then, clone this repo and cd into it.

To deploy to Kubernetes first create a namespace:

kubectl create ns llama

Then apply the manifests under the /deploy/kubernetes directory with

kubectl apply -k deploy/kubernetes/. -n llama

Expose your service however you would normally do that.

OpenAI compatible API

Thanks to llama-cpp-python, a drop-in replacement for OpenAI API is available at http://localhost:3001. Open http://localhost:3001/docs to see the API documentation.

Benchmarks

We've tested LlamaGPT models on the following hardware with the default system prompt, and user prompt: "How does the universe expand?" at temperature 0 to guarantee deterministic results. Generation speed is averaged over the first 10 generations.

Feel free to add your own benchmarks to this table by opening a pull request.

Nous Hermes Llama 2 7B Chat (GGML q4_0)

Device	Generation speed
M1 Max MacBook Pro (64GB RAM)	54 tokens/sec
GCP c2-standard-16 vCPU (64 GB RAM)	16.7 tokens/sec
Ryzen 5700G 4.4GHz 4c (16 GB RAM)	11.50 tokens/sec
GCP c2-standard-4 vCPU (16 GB RAM)	4.3 tokens/sec
Umbrel Home (16GB RAM)	2.7 tokens/sec
Raspberry Pi 4 (8GB RAM)	0.9 tokens/sec

Nous Hermes Llama 2 13B Chat (GGML q4_0)

Device	Generation speed
M1 Max MacBook Pro (64GB RAM)	20 tokens/sec
GCP c2-standard-16 vCPU (64 GB RAM)	8.6 tokens/sec
GCP c2-standard-4 vCPU (16 GB RAM)	2.2 tokens/sec
Umbrel Home (16GB RAM)	1.5 tokens/sec

Nous Hermes Llama 2 70B Chat (GGML q4_0)

Device	Generation speed
M1 Max MacBook Pro (64GB RAM)	4.8 tokens/sec
GCP e2-standard-16 vCPU (64 GB RAM)	1.75 tokens/sec
GCP c2-standard-16 vCPU (64 GB RAM)	1.62 tokens/sec

Code Llama 7B Chat (GGUF Q4_K_M)

Device	Generation speed
M1 Max MacBook Pro (64GB RAM)	41 tokens/sec

Code Llama 13B Chat (GGUF Q4_K_M)

Device	Generation speed
M1 Max MacBook Pro (64GB RAM)	25 tokens/sec

Phind Code Llama 34B Chat (GGUF Q4_K_M)

Device	Generation speed
M1 Max MacBook Pro (64GB RAM)	10.26 tokens/sec

Roadmap and contributing

We're looking to add more features to LlamaGPT. You can see the roadmap here. The highest priorities are:

• Moving the model out of the Docker image and into a separate volume.
• Add Metal support for M1/M2 Macs.
• Add support for Code Llama models.
• Add CUDA support for NVIDIA GPUs.
• Add ability to load custom models.
• Allow users to switch between models.

If you're a developer who'd like to help with any of these, please open an issue to discuss the best way to tackle the challenge. If you're looking to help but not sure where to begin, check out these issues that have specifically been marked as being friendly to new contributors.

Acknowledgements

A massive thank you to the following developers and teams for making LlamaGPT possible:

• Mckay Wrigley for building Chatbot UI.
• Georgi Gerganov for implementing llama.cpp.
• Andrei for building the Python bindings for llama.cpp.
• NousResearch for fine-tuning the Llama 2 7B and 13B models.
• Phind for fine-tuning the Code Llama 34B model.
• Tom Jobbins for quantizing the Llama 2 models.
• Meta for releasing Llama 2 and Code Llama under a permissive license.

---

umbrel.com