Skip to content

Latest commit

 

History

History
156 lines (107 loc) · 5.09 KB

inference-with-fastertransformer.md

File metadata and controls

156 lines (107 loc) · 5.09 KB
Raw

Inference with FasterTransformer

FasterTransformer provides a script and recipe to run the highly optimized transformer-based encoder and decoder component, and it is tested and maintained by NVIDIA.

We adapted the GLM-130B based on Fastertransformer for fast inference, with details in benchmark section.

Download the Model

See Get Model.

Recommend: Run With Docker

Use Docker to quickly build a Flask API application for GLM-130B.

Requirements

Build Container Image

git clone https://github.com/THUDM/FasterTransformer.git
cd FasterTransformer
bash docker/build.sh

Run API With Checkpoints

Set MPSIZE to the number of gpus needed for the checkpoints, and DATA_TYPE to checkpoints precision. The checkpoint we distribute is in 8-way tensor parallel in FP16 precision, a conversion scripts is also provided if you need to change the tensor parallel dimension and the weight precision.

# Convert the checkpoint to MP=4, DATA_TYPE=INT4
python tools/convert_tp.py \
    --input-folder <SRC_CKPT_PATH>  \
    --output-folder <DST_CKPT_PATH> \
    --target-tp 8 \
    --quantization-bit-width 4 \
# Run API
docker run -it --rm --gpus all --shm-size=10g -p 5000:5000 \
           -v <DST_CKPT_PATH>/49300:/checkpoints:ro \
           -e MPSIZE=4 -e DATA_TYPE=int4 \
           ftglm:latest

Test

Benchmark

python3 examples/pytorch/glm/glm_server_test.py

Web Demo

pip install gradio
python3 examples/pytorch/glm/glm_server_frontend_test.py

Manual Configuration

Requirements

  • CMake >= 3.13 for PyTorch
  • CUDA 11.0 or newer version
  • NCCL 2.10 or newer version
  • Python 3 is recommended because some features are not supported in python 2
  • PyTorch: Verify on 1.10.1, >= 1.8.0 should work.

Setup Using Docker

docker run -it --rm --gpus all nvcr.io/nvidia/pytorch:22.09-py3 /bin/bash
conda install -y pybind11

Setup Using Conda

As another way, all the packages can be installed using conda.

Some of our current structure requires that g++ and libtorch produce the same results, so a pre-compiled libtorch may only work with g++-7 or g++-9. And although GLM-130B itself does not rely on openmpi, FasterTransformer requires it during the build process. We are working on these issues.

conda install -y cmake pybind11
conda install -y -c conda-forge cudatoolkit-dev cudnn
cp -r $CONDA_PREFIX/lib/libcudnn* /usr/local/cuda/lib64/
cp -r $CONDA_PREFIX/include/cudnn*.h /usr/local/cuda/include/

If it's hard to install cudatoolkit-dev and cudnn by conda, just install them from NVIDIA Developer, and make sure cmake is able to find cudnn.

cp cudnn/include/cudnn*.h /usr/local/cuda/include
cp cudnn/lib/libcudnn* /usr/local/cuda/lib64
chmod a+r /usr/local/cuda/include/cudnn*.h 
chmod a+r /usr/local/cuda/lib64/libcudnn*

GLM-130B is trained with FP16 precision, a total of 260G of GPU memory is required to store model weights. The model is tested with 8 * 40G A100s.

Build

Get the code and install all dependencies:

git clone https://github.com/THUDM/FasterTransformer.git
mkdir -p FasterTransformer/build
cd FasterTransformer/build
pip3 install icetk transformers

Note: the xx of -DSM=xx in following scripts means the compute capability of your GPU. For example, 60 (P40) or 61 (P4) or 70 (V100) or 75(T4) or 80 (A100) or 86(RTX 3090). Default setting is including 70, 75, 80 and 86.

cmake -DSM=80 -DCMAKE_BUILD_TYPE=Release -DBUILD_PYT=ON -DBUILD_MULTI_GPU=ON ..
make -j

Run GLM-130B

Generate the gemm_config.in file.

# ./bin/gpt_gemm <batch_size> <beam_width> <max_input_len> <head_number> <size_per_head> <inter_size> <vocab_size> <data_type> <tensor_para_size>
./bin/gpt_gemm 1 1 128 96 128 49152 150528 1 8

Running GLM_130B in Pytorch and Flask.

bash ../examples/pytorch/glm/glm-server.sh

You need to check and edit this file to set arguments such as CHECKPOINT_PATH.

Optimization methods

Optimization in GLM_130B are similar to optimization in GPT and GPT-J, describing in the FasterTransformer/gpt_guide.md. Meanwhile, some of the operators are differ from GPT, such as the implementation of RotaryEmbedding, and the use of GeGLU, so we add them additionally into FasterTransformer.

Benchmark

  • Hardware: DGX-A100(8 * 40G)

Encode

Sequence Len 512 1024 2048
Megatron 145 ms 250 ms 453 ms
FasterTransformer 120 ms 220 ms OOM

Decode

Sequence Len 512 1024 2048
Megatron 45.21 s 89.00 s 179.22 s
FasterTransformer 18.77 s 39.81 s 89.88 s