-
Notifications
You must be signed in to change notification settings - Fork 72
Added recipe for Qwen3-235b #145
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
base: main
Are you sure you want to change the base?
Changes from 6 commits
e428e11
47f7068
b07a2df
567597e
261d7fc
ab16df4
0a96978
File filter
Filter by extension
Conversations
Jump to
Diff view
Diff view
There are no files selected for viewing
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,162 @@ | ||
| # Single host inference benchmark of Qwen3-235B with TensorRT-LLM on G4 | ||
|
|
||
| This recipe shows how to serve and benchmark the Qwen3-235B model using [NVIDIA TensorRT-LLM](https://github.com/NVIDIA/TensorRT-LLM) on a single GCP VM with G4 GPUs. For more information on G4 machine types, see the [GCP documentation](https://cloud.google.com/compute/docs/accelerator-optimized-machines#g4-machine-types). | ||
|
|
||
| ## Before you begin | ||
|
|
||
| ### 1. Create a GCP VM with G4 GPUs | ||
|
|
||
| First, we will create a Google Cloud Platform (GCP) Virtual Machine (VM) that has the necessary GPU resources. | ||
|
|
||
| Make sure you have the following prerequisites: | ||
| * [Google Cloud SDK](https://cloud.google.com/sdk/docs/install) is initialized. | ||
| * You have a project with a GPU quota. See [Request a quota increase](https://cloud.google.com/docs/quota/view-request#requesting_higher_quota). | ||
| * [Enable required APIs](https://console.cloud.google.com/flows/enableapi?apiid=compute.googleapis.com). | ||
|
|
||
| The following commands set up environment variables and create a GCE instance. The `MACHINE_TYPE` is set to `g4-standard-192` for 4 GPU VM. The boot disk is set to 200GB to accommodate the models and dependencies. | ||
|
|
||
| ```bash | ||
| export VM_NAME="${USER}-g4-trtllm-qwen3-235b" | ||
| export PROJECT_ID="your-project-id" | ||
| export ZONE="your-zone" | ||
| export MACHINE_TYPE="g4-standard-192" | ||
| export IMAGE_PROJECT="ubuntu-os-accelerator-images" | ||
| export IMAGE_FAMILY="ubuntu-accelerator-2404-amd64-with-nvidia-570" | ||
|
|
||
| gcloud compute instances create ${VM_NAME} \ | ||
| --machine-type=${MACHINE_TYPE} \ | ||
| --project=${PROJECT_ID} \ | ||
| --zone=${ZONE} \ | ||
| --image-project=${IMAGE_PROJECT} \ | ||
| --image-family=${IMAGE_FAMILY} \ | ||
| --maintenance-policy=TERMINATE \ | ||
| --boot-disk-size=200GB | ||
| ``` | ||
|
|
||
| ### 2. Connect to the VM | ||
|
|
||
| Use `gcloud compute ssh` to connect to the newly created instance. | ||
|
|
||
| ```bash | ||
| gcloud compute ssh ${VM_NAME?} --project=${PROJECT_ID?} --zone=${ZONE?} | ||
| ``` | ||
|
|
||
| ```bash | ||
| # Run NVIDIA smi to verify the driver installation and see the available GPUs. | ||
| nvidia-smi | ||
| ``` | ||
|
|
||
| ## Serve a model | ||
|
|
||
| ### 1. Install Docker | ||
|
|
||
| Before you can serve the model, you need to have Docker installed on your VM. You can follow the official documentation to install Docker on Ubuntu: | ||
| [Install Docker Engine on Ubuntu](https://docs.docker.com/engine/install/ubuntu/) | ||
|
|
||
| After installing Docker, make sure the Docker daemon is running. | ||
|
|
||
| ### 2. Install NVIDIA Container Toolkit | ||
|
|
||
| To enable Docker containers to access the GPU, you need to install the NVIDIA Container Toolkit. | ||
|
|
||
| You can follow the official NVIDIA documentation to install the container toolkit: | ||
| [NVIDIA Container Toolkit Install Guide](https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/latest/install-guide.html) | ||
|
|
||
| ### 3. Setup TensorRT-LLM | ||
|
|
||
| ```bash | ||
| sudo apt-get update | ||
| sudo apt-get -y install git git-lfs | ||
|
|
||
| git clone https://github.com/NVIDIA/TensorRT-LLM.git | ||
| cd TensorRT-LLM | ||
| git checkout v1.2.0rc3 | ||
| git submodule update --init --recursive | ||
| git lfs install | ||
| git lfs pull | ||
|
|
||
| # Build the Docker image | ||
| make -C docker release_build | ||
|
|
||
| # Run the Docker container | ||
| mkdir -p /scratch/cache | ||
| make -C docker release_run DOCKER_RUN_ARGS="-v /scratch:/scratch -v /scratch/cache:/root/.cache --ipc=host" | ||
| ``` | ||
| ### 4. Download the Model | ||
|
|
||
| ```bash | ||
| # Inside the container | ||
|
|
||
| # Download the base model from Hugging Face | ||
| apt-get update && apt-get install -y huggingface-cli | ||
|
|
||
| huggingface-cli download Qwen/Qwen3-235B --local-dir /scratch/models/Qwen3-235B | ||
|
|
||
| ``` | ||
|
|
||
| ### 5. Quantize the model using FP8 | ||
| ```bash | ||
| git clone https://github.com/NVIDIA/TensorRT-Model-Optimizer.git | ||
|
|
||
| pushd TensorRT-Model-Optimizer | ||
|
|
||
| pip install -e . | ||
|
|
||
| python examples/llm_ptq/hf_ptq.py \ | ||
| --pyt_ckpt_path /scratch/models/Qwen3-235B \ | ||
| --qformat fp8 \ | ||
| --export_path /scratch/models/exported_model_qwen3_235b_fp8 \ | ||
| --trust_remote_code | ||
|
|
||
| ``` | ||
| ## Run Benchmarks | ||
|
|
||
| Create a script to run the benchmarks with different configurations. | ||
|
|
||
| ```bash | ||
| # Inside the container | ||
|
|
||
| cat << 'EOF' > /scratch/run_benchmark.sh | ||
| #!/bin/bash | ||
|
|
||
| # Function to run benchmarks | ||
| run_benchmark() { | ||
| local model_name=$1 | ||
| local isl=$2 | ||
| local osl=$3 | ||
| local num_requests=$4 | ||
| local tp_size=$5 | ||
| local pp_size=$6 | ||
| local ep_size=$7 | ||
|
|
||
| echo "Running benchmark for $model_name with ISL=$isl, OSL=$osl, TP=$tp_size, PP=$pp_size, EP=$ep_size" | ||
|
|
||
| dataset_file="/scratch/token-norm-dist_${model_name##*/}_${isl}_${osl}.json" | ||
|
|
||
| python benchmarks/cpp/prepare_dataset.py --tokenizer=$model_name --stdout token-norm-dist --num-requests=$num_requests --input-mean=$isl --output-mean=$osl --input-stdev=0 --output-stdev=0 > $dataset_file | ||
|
|
||
| # Save throughput output to a file | ||
| trtllm-bench --model $model_name --model_path ${model_name} throughput --concurrency 128 --dataset $dataset_file --tp $tp_size --pp $pp_size --ep $ep_size --backend pytorch > "/scratch/output_${model_name##*/}_isl${isl}_osl${osl}_tp${tp_size}_pp${pp_size}_ep${ep_size}_throughput.txt" | ||
|
|
||
| rm -f $dataset_file | ||
| } | ||
|
|
||
| model_name="/scratch/models/exported_model_qwen3_235b_fp8" | ||
| TP_SIZE=1 | ||
| PP_SIZE=1 | ||
| EP_SIZE=1 | ||
|
|
||
|
Collaborator
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. Add the setting for NCCL_P2P_LEVEL and a link to the doc for setting: https://docs.cloud.google.com/compute/docs/accelerator-optimized-machines#g4-gpu-p2p
Author
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. Updated the recipe as suggested |
||
| run_benchmark "$model_name" 128 128 1024 $TP_SIZE $PP_SIZE $EP_SIZE | ||
| run_benchmark "$model_name" 128 2048 1024 $TP_SIZE $PP_SIZE $EP_SIZE | ||
| run_benchmark "$model_name" 128 4096 1024 $TP_SIZE $PP_SIZE $EP_SIZE | ||
| run_benchmark "$model_name" 500 2000 1024 $TP_SIZE $PP_SIZE $EP_SIZE | ||
| run_benchmark "$model_name" 1000 1000 1024 $TP_SIZE $PP_SIZE $EP_SIZE | ||
| run_benchmark "$model_name" 2048 128 1024 $TP_SIZE $PP_SIZE $EP_SIZE | ||
| run_benchmark "$model_name" 2048 2048 1024 $TP_SIZE $PP_SIZE $EP_SIZE | ||
| run_benchmark "$model_name" 5000 500 1024 $TP_SIZE $PP_SIZE $EP_SIZE | ||
| run_benchmark "$model_name" 20000 2000 1024 $TP_SIZE $PP_SIZE $EP_SIZE | ||
| EOF | ||
|
|
||
| chmod +x /scratch/run_benchmark.sh | ||
| /scratch/run_benchmark.sh | ||
|
|
||
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Add an addditional optional section (5.1 Optional) to use a pre-quantized model, such as Nvidia's official FP8 checkpoint: https://huggingface.co/nvidia/Qwen3-235B-A22B-FP8
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Updated the recipe as suggested