Skip to content

Latest commit

 

History

History
 
 

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
 
 
 
 
 
 

Mamba

In this directory, you will find examples on how you could use IPEX-LLM optimize_model API to accelerate Mamba models. For illustration purposes, we utilize the state-spaces/mamba-1.4b and state-spaces/mamba-2.8b as reference Mamba models.

Requirements

To run these examples with IPEX-LLM, we have some recommended requirements for your machine, please refer to here for more information.

Example: Predict Tokens using generate() API

In the example generate.py, we show a basic use case for a Mamba model to predict the next N tokens using generate() API, with IPEX-LLM INT4 optimizations.

1. Install

We suggest using conda to manage the Python environment. For more information about conda installation, please refer to here.

After installing conda, create a Python environment for IPEX-LLM:

On Linux:

conda create -n llm python=3.11 # recommend to use Python 3.11
conda activate llm

# install the latest ipex-llm nightly build with 'all' option
pip install --pre --upgrade ipex-llm[all] --extra-index-url https://download.pytorch.org/whl/cpu
pip install einops # package required by Mamba

On Windows:

conda create -n llm python=3.11
conda activate llm

pip install --pre --upgrade ipex-llm[all]
pip install einops

2. Run

After setting up the Python environment, you could run the example by following steps.

2.1 Client

On client Windows machines, it is recommended to run directly with full utilization of all cores:

python ./generate.py

More information about arguments can be found in Arguments Info section. The expected output can be found in Sample Output section.

2.2 Server

For optimal performance on server, it is recommended to set several environment variables (refer to here for more information), and run the example with all the physical cores of a single socket.

E.g. on Linux,

# set IPEX-LLM env variables
source ipex-llm-init

# e.g. for a server with 48 cores per socket
export OMP_NUM_THREADS=48
numactl -C 0-47 -m 0 python ./generate.py

More information about arguments can be found in Arguments Info section. The expected output can be found in Sample Output section.

2.3 Arguments Info

In the example, several arguments can be passed to satisfy your requirements:

  • --repo-id-or-model-path: str, argument defining the huggingface repo id for the Mamba model (e.g state-spaces/mamba-1.4b and state-spaces/mamba-2.8b) to be downloaded, or the path to the huggingface checkpoint folder. It is default to be state-spaces/mamba-1.4b.
  • --tokenizer-repo-id-or-path: str, argument defining the huggingface repo id for the tokenizer of Mamba model to be downloaded, or the path to the huggingface checkpoint folder. It is default to be EleutherAI/gpt-neox-20b.
  • --prompt: str, argument defining the prompt to be inferred (with integrated prompt format for chat). It is default to be 'What is AI?'.
  • --n-predict: int, argument defining the max number of tokens to predict. It is default to be 32.

2.4 Sample Output

Inference time: xxxx s
-------------------- Output --------------------
What is AI?

Artificial Intelligence is a field of computer science that deals with the creation of machines that can learn and think like humans. It is a field that has
Inference time: xxxx s
-------------------- Output --------------------
What is AI?

Artificial Intelligence is a field of computer science that focuses on developing intelligent machines. It is a field that is concerned with the creation of machines that can