EmbeddedLLM

Run local LLMs on iGPU, APU and CPU (AMD , Intel, and Qualcomm (Coming Soon)). Easiest way to launch OpenAI API Compatible Server on Windows, Linux and MacOS

Support matrix	Supported now	Under Development	On the roadmap
Model architectures	Gemma Llama * Mistral + Phi
Platform	Linux Windows
Architecture	x86 x64	Arm64
Hardware Acceleration	CUDA DirectML IpexLLM	QNN ROCm	OpenVINO

* The Llama model architecture supports similar model families such as CodeLlama, Vicuna, Yi, and more.

+ The Mistral model architecture supports similar model families such as Zephyr.

🚀 Latest News

• [2024/06] Support Phi-3 (mini, small, medium), Phi-3-Vision-Mini, Llama-2, Llama-3, Gemma (v1), Mistral v0.3, Starling-LM, Yi-1.5.
• [2024/06] Support vision/chat inference on iGPU, APU, CPU and CUDA.

Table Content

• Supported Models
  • Onnxruntime Models
  • Ipex-LLM Models
• Getting Started
  • Installation From Source
  • Launch OpenAI API Compatible Server
  • Launch Chatbot Web UI
  • Launch Model Management UI
• Compile OpenAI-API Compatible Server into Windows Executable
• Prebuilt Binary (Alpha)
• Acknowledgements

Supported Models (Quick Start)

Models	Parameters	Context Length	Link
Gemma-2b-Instruct v1	2B	8192	EmbeddedLLM/gemma-2b-it-onnx
Llama-2-7b-chat	7B	4096	EmbeddedLLM/llama-2-7b-chat-int4-onnx-directml
Llama-2-13b-chat	13B	4096	EmbeddedLLM/llama-2-13b-chat-int4-onnx-directml
Llama-3-8b-chat	8B	8192	EmbeddedLLM/mistral-7b-instruct-v0.3-onnx
Mistral-7b-v0.3-instruct	7B	32768	EmbeddedLLM/mistral-7b-instruct-v0.3-onnx
Phi-3-mini-4k-instruct-062024	3.8B	4096	EmbeddedLLM/Phi-3-mini-4k-instruct-062024-onnx
Phi3-mini-4k-instruct	3.8B	4096	microsoft/Phi-3-mini-4k-instruct-onnx
Phi3-mini-128k-instruct	3.8B	128k	microsoft/Phi-3-mini-128k-instruct-onnx
Phi3-medium-4k-instruct	17B	4096	microsoft/Phi-3-medium-4k-instruct-onnx-directml
Phi3-medium-128k-instruct	17B	128k	microsoft/Phi-3-medium-128k-instruct-onnx-directml
Openchat-3.6-8b	8B	8192	EmbeddedLLM/openchat-3.6-8b-20240522-onnx
Yi-1.5-6b-chat	6B	32k	EmbeddedLLM/01-ai_Yi-1.5-6B-Chat-onnx
Phi-3-vision-128k-instruct		128k	EmbeddedLLM/Phi-3-vision-128k-instruct-onnx

Getting Started

Installation

From Source

• Windows

  1. Custom Setup:
  • IPEX(XPU): Requires anaconda environment. conda create -n ellm python=3.10 libuv; conda activate ellm.
  • DirectML: If you are using Conda Environment. Install additional dependencies: conda install conda-forge::vs2015_runtime.

  2. Install embeddedllm package. $env:ELLM_TARGET_DEVICE='directml'; pip install -e .. Note: currently support cpu, directml and cuda.

     • DirectML: $env:ELLM_TARGET_DEVICE='directml'; pip install -e .[directml]
     • CPU: $env:ELLM_TARGET_DEVICE='cpu'; pip install -e .[cpu]
     • CUDA: $env:ELLM_TARGET_DEVICE='cuda'; pip install -e .[cuda]
     • IPEX: $env:ELLM_TARGET_DEVICE='ipex'; python setup.py develop
     • OpenVINO: $env:ELLM_TARGET_DEVICE='openvino'; pip install -e .[openvino]
     • With Web UI:
       • DirectML: $env:ELLM_TARGET_DEVICE='directml'; pip install -e .[directml,webui]
       • CPU: $env:ELLM_TARGET_DEVICE='cpu'; pip install -e .[cpu,webui]
       • CUDA: $env:ELLM_TARGET_DEVICE='cuda'; pip install -e .[cuda,webui]
       • IPEX: $env:ELLM_TARGET_DEVICE='ipex'; python setup.py develop; pip install -r requirements-webui.txt
       • OpenVINO: $env:ELLM_TARGET_DEVICE='openvino'; pip install -e .[openvino,webui]

• Linux

  1. Custom Setup:
  • IPEX(XPU): Requires anaconda environment. conda create -n ellm python=3.10 libuv; conda activate ellm.
  • DirectML: If you are using Conda Environment. Install additional dependencies: conda install conda-forge::vs2015_runtime.

  2. Install embeddedllm package. ELLM_TARGET_DEVICE='directml' pip install -e .. Note: currently support cpu, directml and cuda.

     • DirectML: ELLM_TARGET_DEVICE='directml' pip install -e .[directml]
     • CPU: ELLM_TARGET_DEVICE='cpu' pip install -e .[cpu]
     • CUDA: ELLM_TARGET_DEVICE='cuda' pip install -e .[cuda]
     • IPEX: ELLM_TARGET_DEVICE='ipex' python setup.py develop
     • OpenVINO: ELLM_TARGET_DEVICE='openvino' pip install -e .[openvino]
     • With Web UI:
       • DirectML: ELLM_TARGET_DEVICE='directml' pip install -e .[directml,webui]
       • CPU: ELLM_TARGET_DEVICE='cpu' pip install -e .[cpu,webui]
       • CUDA: ELLM_TARGET_DEVICE='cuda' pip install -e .[cuda,webui]
       • IPEX: ELLM_TARGET_DEVICE='ipex' python setup.py develop; pip install -r requirements-webui.txt
       • OpenVINO: ELLM_TARGET_DEVICE='openvino' pip install -e .[openvino,webui]

Launch OpenAI API Compatible Server

1. Custom Setup:

   • Ipex

     • For Intel iGPU:

       set SYCL_CACHE_PERSISTENT=1
       set BIGDL_LLM_XMX_DISABLED=1

     • For Intel Arc™ A-Series Graphics:

       set SYCL_CACHE_PERSISTENT=1

2. ellm_server --model_path <path/to/model/weight>.

3. Example code to connect to the api server can be found in scripts/python. Note: To find out more of the supported arguments. ellm_server --help.

Launch Chatbot Web UI

1. ellm_chatbot --port 7788 --host localhost --server_port <ellm_server_port> --server_host localhost. Note: To find out more of the supported arguments. ellm_chatbot --help.

Launch Model Management UI

It is an interface that allows you to download and deploy OpenAI API compatible server. You can find out the disk space required to download the model in the UI.

1. ellm_modelui --port 6678. Note: To find out more of the supported arguments. ellm_modelui --help.

Compile OpenAI-API Compatible Server into Windows Executable

NOTE: OpenVINO packaging currently uses torch==2.4.0. It will not be able to run due to missing dependencies which is libomp. Make sure to install libomp and add the libomp-xxxxxxx.dll to C:\\Windows\\System32.

1. Install embeddedllm.

2. Install PyInstaller: pip install pyinstaller==6.9.0.

3. Compile Windows Executable: pyinstaller .\ellm_api_server.spec.

4. You can find the executable in the dist\ellm_api_server.

5. Use it like ellm_server. .\ellm_api_server.exe --model_path <path/to/model/weight>.

   Powershell/Terminal Usage:

   ellm_server --model_path <path/to/model/weight>
   
   # DirectML
   ellm_server --model_path 'EmbeddedLLM/Phi-3-mini-4k-instruct-onnx-directml' --port 5555
   
   # IPEX-LLM
   ellm_server --model_path '.\meta-llama_Meta-Llama-3.1-8B-Instruct\'  --backend 'ipex' --device 'xpu' --port 5555 --served_model_name 'meta-llama_Meta/Llama-3.1-8B-Instruct'
   
   # OpenVINO
   ellm_server --model_path '.\meta-llama_Meta-Llama-3.1-8B-Instruct\'  --backend 'openvino' --device 'gpu' --port 5555 --served_model_name 'meta-llama_Meta/Llama-3.1-8B-Instruct'

Prebuilt OpenAI API Compatible Windows Executable (Alpha)

You can find the prebuilt OpenAI API Compatible Windows Executable in the Release page.

Powershell/Terminal Usage (Use it like ellm_server):

.\ellm_api_server.exe --model_path <path/to/model/weight>

# DirectML
.\ellm_api_server.exe --model_path 'EmbeddedLLM_Phi-3-mini-4k-instruct-062024-onnx\onnx\directml\Phi-3-mini-4k-instruct-062024-int4' --port 5555

# IPEX-LLM
.\ellm_api_server.exe --model_path '.\meta-llama_Meta-Llama-3.1-8B-Instruct\'  --backend 'ipex' --device 'xpu' --port 5555 --served_model_name 'meta-llama_Meta/Llama-3.1-8B-Instruct'

# OpenVINO
.\ellm_api_server.exe --model_path '.\meta-llama_Meta-Llama-3.1-8B-Instruct\'  --backend 'openvino' --device 'gpu' --port 5555 --served_model_name 'meta-llama_Meta/Llama-3.1-8B-Instruct'

Acknowledgements

• Excellent open-source projects: vLLM, onnxruntime-genai, Ipex-LLM and many others.