Skip to content

Latest commit

 

History

History
197 lines (146 loc) · 8.41 KB

README.md

File metadata and controls

197 lines (146 loc) · 8.41 KB
Raw

LlamaIndex Examples

This folder contains examples showcasing how to use LlamaIndex with ipex-llm.

LlamaIndex is a data framework designed to improve large language models by providing tools for easier data ingestion, management, and application integration.

Retrieval-Augmented Generation (RAG) Example

The RAG example (rag.py) is adapted from the Official llama index RAG example. This example builds a pipeline to ingest data (e.g. llama2 paper in pdf format) into a vector database (e.g. PostgreSQL), and then build a retrieval pipeline from that vector database.

1. Install Prerequisites

To benefit from IPEX-LLM on Intel GPUs, there are several prerequisite steps for tools installation and environment preparation.

If you are a Windows user, visit the Install IPEX-LLM on Windows with Intel GPU Guide, and follow Install Prerequisites to update GPU driver (optional) and install Conda.

If you are a Linux user, visit the Install IPEX-LLM on Linux with Intel GPU, and follow Install Prerequisites to install GPU driver, Intel® oneAPI Base Toolkit 2024.0, and Conda.

2. Setting up Dependencies

  • Install LlamaIndex Packages 
    conda activate <your-conda-env-name>
pip install llama-index-llms-ipex-llm[xpu]==0.1.8 --extra-index-url https://pytorch-extension.intel.com/release-whl/stable/xpu/us/
pip install llama-index-embeddings-ipex-llm[xpu]==0.1.5 --extra-index-url https://pytorch-extension.intel.com/release-whl/stable/xpu/us/
pip install llama-index-readers-file==0.1.33
pip install llama-index-vector-stores-postgres==0.1.14
pip install pymupdf

Note

  • You could refer llama-index-llms-ipex-llm and llama-index-embeddings-ipex-llm for more information.
  • The installation of llama-index-llms-ipex-llm or llama-index-embeddings-ipex-llm will also install IPEX-LLM and its dependencies.
  • You can also use https://pytorch-extension.intel.com/release-whl/stable/xpu/cn/ as the extra-indel-url.
  • IpexLLMEmbedding currently only provides optimization for Hugging Face Bge models.

  • Database Setup (using PostgreSQL):

    • Linux

      • Installation:

        sudo apt-get install postgresql-client
sudo apt-get install postgresql

      • Initialization:

        Switch to the postgres user and launch psql console

        sudo su - postgres
psql

        Then, create a new user role:

        CREATE ROLE <user> WITH LOGIN PASSWORD '<password>';
ALTER ROLE <user> SUPERUSER;
    • Windows
      • click Download the installer in PostgreSQL.
      • Run the downloaded installation package as administrator, then click next continuously.
      • Open PowerShell:
          cd C:\Program Files\PostgreSQL\14\bin
      The exact path will vary depending on your PostgreSQL location.
      • Then in PowerShell:
          .\psql -U postgres
      Input the password you set in the previous installation. If PowerShell shows postgres=#, it indicates a successful connection.
      • Create a new user role:
      CREATE ROLE <user> WITH LOGIN PASSWORD '<password>';
ALTER ROLE <user> SUPERUSER;

  • Pgvector Installation:

    • Linux
    • Windows
      • It is recommended to use pgvector for Windows instead of others (such as conda-force) to avoid potential errors. Some steps may require running as administrator.

  • Data Preparation: Download the Llama2 paper and save it as data/llama2.pdf, which serves as the default source file for retrieval.

    mkdir data
wget --user-agent "Mozilla" "https://arxiv.org/pdf/2307.09288.pdf" -O "data/llama2.pdf"

3. Configures OneAPI environment variables for Linux

Note

Skip this step if you are running on Windows.

This is a required step on Linux for APT or offline installed oneAPI. Skip this step for PIP-installed oneAPI.

source /opt/intel/oneapi/setvars.sh

4. Runtime Configurations

For optimal performance, it is recommended to set several environment variables. Please check out the suggestions based on your device.

4.1 Configurations for Linux

For Intel Arc™ A-Series Graphics and Intel Data Center GPU Flex Series 
export USE_XETLA=OFF
export SYCL_PI_LEVEL_ZERO_USE_IMMEDIATE_COMMANDLISTS=1
export SYCL_CACHE_PERSISTENT=1
For Intel Data Center GPU Max Series 
export LD_PRELOAD=${LD_PRELOAD}:${CONDA_PREFIX}/lib/libtcmalloc.so
export SYCL_PI_LEVEL_ZERO_USE_IMMEDIATE_COMMANDLISTS=1
export SYCL_CACHE_PERSISTENT=1
export ENABLE_SDP_FUSION=1

Note: Please note that libtcmalloc.so can be installed by conda install -c conda-forge -y gperftools=2.10.

For Intel iGPU 
export SYCL_CACHE_PERSISTENT=1
export BIGDL_LLM_XMX_DISABLED=1

4.2 Configurations for Windows

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

Note

For the first time that each model runs on Intel iGPU/Intel Arc™ A300-Series or Pro A60, it may take several minutes to compile.

5. Running the RAG example

In the current directory, run the example with command:

python rag.py -m <path_to_model> -t <path_to_tokenizer>

Additional Parameters for Configuration:

  • -m MODEL_PATH: Required, path to the LLM model
  • -e EMBEDDING_MODEL_PATH: path to the embedding model
  • -u USERNAME: username in the PostgreSQL database
  • -p PASSWORD: password in the PostgreSQL database
  • -q QUESTION: question you want to ask
  • -d DATA: path to source data used for retrieval (in pdf format)
  • -n N_PREDICT: max predict tokens
  • -t TOKENIZER_PATH: Required, path to the tokenizer model

6. Example Output

A query such as "How does Llama 2 compare to other open-source models?" with the Llama2 paper as the data source, using the Llama-2-7b-chat-hf model, will produce the output like below:

The comparison between Llama 2 and other open-source models is complex and depends on various factors such as the specific benchmarks used, the model size, and the task at hand.

In terms of performance on the benchmarks provided in the table, Llama 2 outperforms other open-source models on most categories. For example, on the MMLU benchmark, Llama 2 achieves a score of 22.5, while the next best open-source model, Poplar Aggregated Benchmarks, scores 17.5. Similarly, on the BBH benchmark, Llama 2 scores 20.5, while the next best open-source model scores 16.5.

However, it's important to note that the performance of Llama 2 can vary depending on the specific task and dataset being used. For example, on the coding benchmarks, Llama 2 performs significantly worse than other open-source models, such as PaLM (540B) and GPT-4.

In conclusion, while Llama 2 performs well on most benchmarks compared to other open-source models, its performance

7. Trouble shooting

7.1 Core dump

If you encounter a core dump error in your Python code, it is crucial to verify that the import torch statement is placed at the top of your Python file, just as what we did in rag.py.