tgi-multi-lora-deployment

title	type
Deploy Gemma2 with multiple LoRA adapters with TGI DLC on GKE	inference

Deploy Gemma2 with multiple LoRA adapters with TGI DLC on GKE

Gemma 2 is an advanced, lightweight open model that enhances performance and efficiency while building on the research and technology of its predecessor and the Gemini models developed by Google DeepMind and other teams across Google. Text Generation Inference (TGI) is a toolkit developed by Hugging Face for deploying and serving LLMs, with high performance text generation. And, Google Kubernetes Engine (GKE) is a fully-managed Kubernetes service in Google Cloud that can be used to deploy and operate containerized applications at scale using GCP's infrastructure.

This example showcases how to deploy Gemma 2 2B from the Hugging Face Hub with multiple LoRA adapters fine-tuned for different purposes such as coding, SQL, or Japanese, on a GKE Cluster running the Hugging Face DLC for TGI i.e. a purpose-built container to deploy LLMs in a secure and managed environment.

Setup / Configuration

First, you need to install both gcloud and kubectl in your local machine, which are the command-line tools for Google Cloud and Kubernetes, respectively, to interact with the GCP and the GKE Cluster.

• To install gcloud, follow the instructions at Cloud SDK Documentation - Install the gcloud CLI.
• To install kubectl, follow the instructions at Kubernetes Documentation - Install Tools.

Optionally, to ease the usage of the commands within this tutorial, you need to set the following environment variables for GCP:

export PROJECT_ID=your-project-id
export LOCATION=your-location
export CLUSTER_NAME=your-cluster-name

Then you need to login into your GCP account and set the project ID to the one you want to use for the deployment of the GKE Cluster.

gcloud auth login
gcloud auth application-default login  # For local development
gcloud config set project $PROJECT_ID

Once you are logged in, you need to enable the necessary service APIs in GCP i.e. the Google Kubernetes Engine API and the Google Container Registry API, which are necessary for the deployment of the GKE Cluster and the Hugging Face DLC for TGI.

gcloud services enable container.googleapis.com
gcloud services enable containerregistry.googleapis.com

Additionally, to use kubectl with the GKE Cluster credentials, you also need to install the gke-gcloud-auth-plugin, that can be installed with gcloud as follows:

gcloud components install gke-gcloud-auth-plugin

Note

Installing the gke-gcloud-auth-plugin does not need to be installed via gcloud specifically, to read more about the alternative installation methods, please visit GKE Documentation - Install kubectl and configure cluster access.

Create GKE Cluster

Once everything's set up, you can proceed with the creation of the GKE Cluster and the node pool, which in this case will be a single GPU node, in order to use the GPU accelerator for high performance inference, also following TGI recommendations based on their internal optimizations for GPUs.

To deploy the GKE Cluster, the "Autopilot" mode will be used as it is the recommended one for most of the workloads, since the underlying infrastructure is managed by Google. Alternatively, you can also use the "Standard" mode.

Note

Important to check before creating the GKE Autopilot Cluster the GKE Documentation - Optimize Autopilot Pod performance by choosing a machine series, since not all the versions support GPU accelerators e.g. nvidia-l4 is not supported in the GKE cluster versions 1.28.3 or lower.

gcloud container clusters create-auto $CLUSTER_NAME \
    --project=$PROJECT_ID \
    --location=$LOCATION \
    --release-channel=stable \
    --cluster-version=1.29 \
    --no-autoprovisioning-enable-insecure-kubelet-readonly-port

Note

To select the specific version in your location of the GKE Cluster, you can run the following command:

gcloud container get-server-config \
    --flatten="channels" \
    --filter="channels.channel=STABLE" \
    --format="yaml(channels.channel,channels.defaultVersion)" \
    --location=$LOCATION

For more information please visit GKE Documentation - Specifying cluster version.

Once the GKE Cluster is created, you can get the credentials to access it via kubectl with the following command:

gcloud container clusters get-credentials $CLUSTER_NAME --location=$LOCATION

Get Hugging Face token and set secrets in GKE

As google/gemma-2-2b-it is a gated model, you need to set a Kubernetes secret with the Hugging Face Hub token via kubectl.

To generate a custom token for the Hugging Face Hub, you can follow the instructions at Hugging Face Hub - User access tokens; and the recommended way of setting it is to install the huggingface_hub Python SDK as follows:

pip install --upgrade --quiet huggingface_hub

And then login in with the generated token with read-access over the gated/private model:

huggingface-cli login

Finally, you can create the Kubernetes secret with the generated token for the Hugging Face Hub as follows using the huggingface_hub Python SDK to retrieve the token:

kubectl create secret generic hf-secret \
    --from-literal=hf_token=$(python -c "from huggingface_hub import get_token; print(get_token())") \
    --dry-run=client -o yaml | kubectl apply -f -

Or, alternatively, you can directly set the token as follows:

kubectl create secret generic hf-secret \
    --from-literal=hf_token=hf_*** \
    --dry-run=client -o yaml | kubectl apply -f -

More information on how to set Kubernetes secrets in a GKE Cluster at Secret Manager Documentation - Use Secret Manager add-on with Google Kubernetes Engine.

Deploy TGI

Now you can proceed to the Kubernetes deployment of the Hugging Face DLC for TGI, serving the google/gemma-2-2b-it model and multiple LoRA adapters fine-tuned on top of it, from the Hugging Face Hub.

Note

To explore all the models that can be served via TGI, you can explore the models tagged with text-generation-inference in the Hub.

The Hugging Face DLC for TGI will be deployed via kubectl, from the configuration files in the config/ directory:

• deployment.yaml: contains the deployment details of the pod including the reference to the Hugging Face DLC for TGI setting the MODEL_ID to google/gemma-2-2b-it, and the LORA_ADAPTERS to google-cloud-partnership/gemma-2-2b-it-lora-magicoder,google-cloud-partnership/gemma-2-2b-it-lora-sql, being the following adapters:
  • google-cloud-partnership/gemma-2-2b-it-lora-sql: fine-tuned with gretelai/synthetic_text_to_sql to generate SQL queries with an explanation, given an SQL context and a prompt / question about it.
  • google-cloud-partnership/gemma-2-2b-it-lora-magicoder: fine-tuned with ise-uiuc/Magicoder-OSS-Instruct-75K to generate code in diverse programming languages such as Python, Rust, or C, among many others; based on an input problem.
  • google-cloud-partnership/gemma-2-2b-it-lora-jap-en: fine-tuned with Jofthomas/japanese-english-translation, a synthetically generated dataset of short Japanese sentences translated to English; to translate English to Japanese and the other way around.
• service.yaml: contains the service details of the pod, exposing the port 8080 for the TGI service.
• (optional) ingress.yaml: contains the ingress details of the pod, exposing the service to the external world so that it can be accessed via the ingress IP.

Warning

Note that the selected LoRA adapters are not intended to be used on production environments, as the fine-tuned adapters have not been tested extensively.

git clone https://github.com/huggingface/Google-Cloud-Containers
kubectl apply -f Google-Cloud-Containers/examples/gke/tgi-multi-lora-deployment/config

Note

The Kubernetes deployment may take a few minutes to be ready, so you can check the status of the deployment with the following command:

kubectl get pods

Alternatively, you can just wait for the deployment to be ready with the following command:

kubectl wait --for=condition=Available --timeout=700s deployment/tgi-deployment

Inference with TGI

To run the inference over the deployed TGI service, you need to make sure that the service is accessible first, you can do so by either:

• Port-forwarding the deployed TGI service to the port 8080, so as to access via localhost with the command:

  kubectl port-forward service/tgi-service 8080:8080

• Accessing the TGI service via the external IP of the ingress, which is the default scenario here since you have defined the ingress configuration in the config/ingress.yaml file (but it can be skipped in favour of the port-forwarding), that can be retrieved with the following command:

  kubectl get ingress tgi-ingress -o jsonpath='{.status.loadBalancer.ingress[0].ip}'

Via cURL

To send a POST request to the TGI service using cURL, you can run the following command:

curl http://localhost:8080/v1/chat/completions \
    -X POST \
    -d '{"messages":[{"role":"user","content":"What is Deep Learning?"}],"temperature":0.7,"top_p":0.95,"max_tokens":128}}' \
    -H 'Content-Type: application/json'

Or send a POST request to the ingress IP instead:

curl http://$(kubectl get ingress tgi-ingress -o jsonpath='{.status.loadBalancer.ingress[0].ip}')/v1/chat/completions \
    -X POST \
    -d '{"messages":[{"role":"user","content":"What is Deep Learning?"}],"temperature":0.7,"top_p":0.95,"max_tokens":128}}' \
    -H 'Content-Type: application/json'

Note

As in this case you are serving multiple LoRA adapters, to use those you will need to specify the model parameter when using the /v1/chat/completions endpoint (or the adapter_id parameter when using the /generate endpoint), so that the LoRA adapter is used. In any other case, the base model will be used instead, meaning that the adapters are only used when explicitly specified.

For example, say that you want to generate a piece of code for a problem that you cannot solve, then you should ideally use the fine-tuned adapter google-cloud-partnership/gemma-2-2b-it-lora-magicoder which is specifically fine-tuned for that; alternatively you could also use the base instruction-tuned model as it may be able to tackle a wide variety of tasks, but e.g. the Japanese to English model wouldn't be a nice pick for that task.

curl http://localhost:8080/v1/chat/completions \
    -X POST \
    -d '{"messages":[{"role":"user","content":"You are given a vector of integers, A, of length n. Your task is to implement a function that finds the maximum product of any two distinct elements in the vector. Write a function in Rust to return this maximum product. Function Signature: rust fn max_product(a: Vec<i32>) -> i32  Input: - A vector a of length n (2 <= n <= 10^5), where each element is an integer (-10^4 <= a[i] <= 10^4). Output: - Return the maximum product of two distinct elements. Example: Input: a = vec![1, 5, 3, 9] Output: max_product(a) -> 45"}],"temperature":0.7,"top_p":0.95,"max_tokens":256,"model":"google-cloud-partnership/gemma-2-2b-it-lora-magicoder"}}' \
    -H 'Content-Type: application/json'

Which generates the following solution to the given prompt:

{"object":"chat.completion","id":"","created":1727378101,"model":"google/gemma-2-2b-it","system_fingerprint":"2.3.1-dev0-native","choices":[{"index":0,"message":{"role":"assistant","content":"\`\`\`rust\nfn max_product(a: Vec<i32>) -> i32 {\n    let mut max1 = a[0];\n    let mut max2 = a[1];\n    if max2 < max1 {\n        std::mem::swap(&mut max1, &mut max2);\n    }\n    for i in 2..a.len() {\n        if a[i] > max1 {\n            max2 = max1;\n            max1 = a[i];\n        } else if a[i] > max2 {\n            "},"logprobs":null,"finish_reason":"length"}],"usage":{"prompt_tokens":163,"completion_tokens":128,"total_tokens":291}}

Translated to Rust code that would be:

fn max_product(a: Vec<i32>) -> i32 {
    if a.len() < 2 {
        return 0;
    }
    let mut max_product = a[0] * a[1];
    for i in 1..a.len() {
        for j in i + 1..a.len() {
            if a[i] * a[j] > max_product {
                max_product = a[i] * a[j];
            }
        }
    }
    max_product
}

Via Python

To run the inference using Python, you can either use the huggingface_hub Python SDK (recommended) or the openai Python SDK.

Note

In the examples below localhost will be used, but if you did deploy TGI with the ingress, feel free to use the ingress IP as mentioned above (without specifying the port).

huggingface_hub

You can install it via pip as pip install --upgrade --quiet huggingface_hub, and then run the following snippet to mimic the cURL commands above i.e. sending requests to the Messages API providing the adapter identifier via the model parameter:

from huggingface_hub import InferenceClient

client = InferenceClient(base_url="http://localhost:8080", api_key="-")

chat_completion = client.chat.completions.create(
    model="google-cloud-partnership/gemma-2-2b-it-lora-magicoder",
    messages=[
        {"role": "user", "content": "You are given a vector of integers, A, of length n. Your task is to implement a function that finds the maximum product of any two distinct elements in the vector. Write a function in Rust to return this maximum product. Function Signature: rust fn max_product(a: Vec<i32>) -> i32  Input: - A vector a of length n (2 <= n <= 10^5), where each element is an integer (-10^4 <= a[i] <= 10^4). Output: - Return the maximum product of two distinct elements. Example: Input: a = vec![1, 5, 3, 9] Output: max_product(a) -> 45"},
    ],
    max_tokens=128,
)

Alternatively, you can also format the prompt yourself and send that via the Text Generation API providing the adapter identifier via the adapter_id argument as follows:

from huggingface_hub import InferenceClient

client = InferenceClient("http://localhost:8080", api_key="-")

generation = client.text_generation(
    prompt="You are given a vector of integers, A, of length n. Your task is to implement a function that finds the maximum product of any two distinct elements in the vector. Write a function in Rust to return this maximum product. Function Signature: rust fn max_product(a: Vec<i32>) -> i32  Input: - A vector a of length n (2 <= n <= 10^5), where each element is an integer (-10^4 <= a[i] <= 10^4). Output: - Return the maximum product of two distinct elements. Example: Input: a = vec![1, 5, 3, 9] Output: max_product(a) -> 45",
    max_new_tokens=128,
    adapter_id="google-cloud-partnership/gemma-2-2b-it-lora-magicoder",
)

openai

Additionally, you can also use the Messages API via openai; you can install it via pip as pip install --upgrade openai, and then run:

from openai import OpenAI

client = OpenAI(
    base_url="http://localhost:8080/v1/",
    api_key="-",
)

chat_completion = client.chat.completions.create(
    model="google-cloud-partnership/gemma-2-2b-it-lora-magicoder",
    messages=[
        {"role": "user", "content": "You are given a vector of integers, A, of length n. Your task is to implement a function that finds the maximum product of any two distinct elements in the vector. Write a function in Rust to return this maximum product. Function Signature: rust fn max_product(a: Vec<i32>) -> i32  Input: - A vector a of length n (2 <= n <= 10^5), where each element is an integer (-10^4 <= a[i] <= 10^4). Output: - Return the maximum product of two distinct elements. Example: Input: a = vec![1, 5, 3, 9] Output: max_product(a) -> 45"},
    ],
    max_tokens=128,
)

Delete GKE Cluster

Finally, once you are done using TGI on the GKE Cluster, you can safely delete the GKE Cluster to avoid incurring in unnecessary costs.

gcloud container clusters delete $CLUSTER_NAME --location=$LOCATION

Alternatively, you can also downscale the replicas of the deployed pod to 0 in case you want to preserve the cluster, since the default GKE Cluster deployed with GKE Autopilot mode is running just a single e2-small instance.

kubectl scale --replicas=0 deployment/tgi-deployment