ziya

Ziya

In this directory, you will find examples on how you could apply BigDL-LLM INT4 optimizations on Ziya models. For illustration purposes, we utilize the IDEA-CCNL/Ziya-Coding-34B-v1.0 as a reference Ziya model.

Note: If you want to download the Hugging Face Transformers model, please refer to here.

BigDL-LLM optimizes the Transformers model in INT4 precision at runtime, and thus no explicit conversion is needed.

Requirements

To run these examples with BigDL-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 Ziya model to predict the next N tokens using generate() API, with BigDL-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 BigDL-LLM:

conda create -n llm python=3.9 # recommend to use Python 3.9
conda activate llm

pip install --pre --upgrade bigdl-llm[all] # install the latest bigdl-llm nightly build with 'all' option
pip install einops  # additional package required for Ziya to conduct generation

2. Run

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

Note: When loading the model in 4-bit, BigDL-LLM converts linear layers in the model into INT4 format. In theory, a XB model saved in 16-bit will requires approximately 2X GB of memory for loading, and ~0.5X GB memory for further inference.

Please select the appropriate size of the Ziya model based on the capabilities of your machine.

2.1 Client

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

python ./generate.py --prompt 'def quick_sort(arr):\n'

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 BigDL-LLM env variables
source bigdl-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 --prompt 'def quick_sort(arr):\n'

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 Ziya model to be downloaded, or the path to the huggingface checkpoint folder. It is default to be 'IDEA-CCNL/Ziya-Coding-34B-v1.0'.
• --prompt: str, argument defining the prompt to be inferred (with integrated prompt format for chat). It is default to be def quick_sort(arr):\n.
• --n-predict: int, argument defining the max number of tokens to predict. It is default to be 128.

2.4 Sample Output

IDEA-CCNL/Ziya-Coding-34B-v1.0

Inference time: xxxx s
-------------------- Prompt --------------------
<human>: 
def quick_sort(arr):\n
<bot>: 

-------------------- Output --------------------
<s> <human>: 
def quick_sort(arr):\n
<bot>: 
def partition(arr, low, high):

    i = (low-1)
    pivot = arr[high]
    for j in range(low, high):
        if arr[j] <= pivot:
            arr[i], arr[j] = arr[j], arr[i]
            i = i+1
    arr[i], arr[high] = arr[high], arr[i]
    return i

def quick_sort(arr, low, high):
    if low < high:
        pi = partition(arr, low,