google-cloud/pr_98/en/_app/immutable/nodes/8.fc26b341.js

import{s as Ul,o as xl,n as J}from"../chunks/scheduler.8b74b908.js";import{S as jl,i as kl,g as r,s as i,r as u,A as El,h as c,f as l,c as o,j as He,u as f,x as p,k as vl,y as G,a as n,v as d,d as m,t as g,w as h}from"../chunks/index.0ed2a570.js";import{T as I}from"../chunks/Tip.07d3ac1e.js";import{C as M}from"../chunks/CodeBlock.530268b7.js";import{H as Le,E as Gl}from"../chunks/EditOnGithub.d2d81eda.js";function Il(y){let s,$='Installing the <code>gke-gcloud-auth-plugin</code> does not need to be installed via <code>gcloud</code> specifically, to read more about the alternative installation methods, please visit <a href="https://cloud.google.com/kubernetes-engine/docs/how-to/cluster-access-for-kubectl#install_plugin" rel="nofollow">https://cloud.google.com/kubernetes-engine/docs/how-to/cluster-access-for-kubectl#install_plugin</a>.';return{c(){s=r("p"),s.innerHTML=$},l(a){s=c(a,"P",{"data-svelte-h":!0}),p(s)!=="svelte-sm1xw4"&&(s.innerHTML=$)},m(a,b){n(a,s,b)},p:J,d(a){a&&l(s)}}}function Jl(y){let s,$='CPU is being used to run the inference on top of the text embeddings models to showcase the current capabilities of TEI, but switching to GPU is as easy as replacing <code>spec.containers[0].image</code> with <code>us-docker.pkg.dev/deeplearning-platform-release/gcr.io/huggingface-text-embeddings-inference-cu122.1-4.ubuntu2204</code>, and then updating the requested resources, as well as the <code>nodeSelector</code> requirements in the <code>deployment.yaml</code> file. For more information, please refer to the <a href="https://github.com/huggingface/Google-Cloud-Containers/tree/main/examples/gke/tei-deployment/gpu-config/" rel="nofollow"><code>gpu-config</code></a> directory that contains a pre-defined configuration for GPU serving in TEI with an NVIDIA Tesla T4 GPU (with a compute capability of 7.5 i.e. natively supported in TEI).';return{c(){s=r("p"),s.innerHTML=$},l(a){s=c(a,"P",{"data-svelte-h":!0}),p(s)!=="svelte-11sq884"&&(s.innerHTML=$)},m(a,b){n(a,s,b)},p:J,d(a){a&&l(s)}}}function _l(y){let s,$='Important to check before creating the GKE Autopilot Cluster the <a href="https://cloud.google.com/kubernetes-engine/docs/how-to/performance-pods" rel="nofollow">GKE Documentation - Optimize Autopilot Pod performance by choosing a machine series</a>, since not all the cluster versions support every CPU. Same applies for the GPU support e.g. <code>nvidia-l4</code> is not supported in the GKE cluster versions 1.28.3 or lower.';return{c(){s=r("p"),s.innerHTML=$},l(a){s=c(a,"P",{"data-svelte-h":!0}),p(s)!=="svelte-1nupym3"&&(s.innerHTML=$)},m(a,b){n(a,s,b)},p:J,d(a){a&&l(s)}}}function Ll(y){let s,$="To select the specific version in your location of the GKE Cluster, you can run the following command:";return{c(){s=r("p"),s.textContent=$},l(a){s=c(a,"P",{"data-svelte-h":!0}),p(s)!=="svelte-a3yff9"&&(s.textContent=$)},m(a,b){n(a,s,b)},p:J,d(a){a&&l(s)}}}function Zl(y){let s,$='Recently, the Hugging Face Hub team has included the <code>text-embeddings-inference</code> tag in the Hub, so feel free to explore all the embedding models in the Hub that can be served via TEI at <a href="https://huggingface.co/models?other=text-embeddings-inference" rel="nofollow">https://huggingface.co/models?other=text-embeddings-inference</a>.';return{c(){s=r("p"),s.innerHTML=$},l(a){s=c(a,"P",{"data-svelte-h":!0}),p(s)!=="svelte-g6mrbh"&&(s.innerHTML=$)},m(a,b){n(a,s,b)},p:J,d(a){a&&l(s)}}}function Vl(y){let s,$='As already mentioned, for this example you will be deploying the container in a CPU node, but the configuration to deploy TEI in a GPU node is also available in the <a href="https://github.com/huggingface/Google-Cloud-Containers/tree/main/examples/gke/tei-deployment/gpu-config/" rel="nofollow"><code>gpu-config</code></a> directory, so if you want to deploy TEI in a GPU node, please run <code>kubectl apply -f gpu-config/</code> instead of <code>kubectl apply -f cpu-config/</code>.';return{c(){s=r("p"),s.innerHTML=$},l(a){s=c(a,"P",{"data-svelte-h":!0}),p(s)!=="svelte-1aq8fgr"&&(s.innerHTML=$)},m(a,b){n(a,s,b)},p:J,d(a){a&&l(s)}}}function Nl(y){let s,$="The Kubernetes deployment may take a few minutes to be ready, so you can check the status of the deployment with the following command:";return{c(){s=r("p"),s.textContent=$},l(a){s=c(a,"P",{"data-svelte-h":!0}),p(s)!=="svelte-qgh43p"&&(s.textContent=$)},m(a,b){n(a,s,b)},p:J,d(a){a&&l(s)}}}function Hl(y){let s,$="TEI exposes different inference endpoints based on the task that the model is serving:";return{c(){s=r("p"),s.textContent=$},l(a){s=c(a,"P",{"data-svelte-h":!0}),p(s)!=="svelte-2zu4d0"&&(s.textContent=$)},m(a,b){n(a,s,b)},p:J,d(a){a&&l(s)}}}function Al(y){let s,$,a,b,_,Ae,L,Yt="Snowflake’s Arctic Embed is a suite of text embedding models that focuses on creating high-quality retrieval models optimized for performance, achieving state-of-the-art (SOTA) performance on the MTEB/BEIR leaderboard for each of their size variants. Text Embeddings Inference (TEI) is a toolkit developed by Hugging Face for deploying and serving open source text embeddings and sequence classification models; enabling high-performance extraction for the most popular models, including FlagEmbedding, Ember, GTE and E5. 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.",Pe,Z,Bt="This example showcases how to deploy a text embedding model from the Hugging Face Hub on a GKE Cluster running a purpose-built container to deploy text embedding models in a secure and managed environment with the Hugging Face DLC for TEI.",We,V,Fe,N,Dt="First, you need to install both <code>gcloud</code> and <code>kubectl</code> 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.",Qe,H,Xt='<li>To install <code>gcloud</code>, follow the instructions at <a href="https://cloud.google.com/sdk/docs/install" rel="nofollow">Cloud SDK Documentation - Install the gcloud CLI</a>.</li> <li>To install <code>kubectl</code>, follow the instructions at <a href="https://kubernetes.io/docs/tasks/tools/#kubectl" rel="nofollow">Kubernetes Documentation - Install Tools</a>.</li>',Se,A,qt="Optionally, to ease the usage of the commands within this tutorial, you need to set the following environment variables for GCP:",Re,P,Ye,W,zt="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.",Be,F,De,Q,Ot="Once you are logged in, you need to enable the necessary service APIs in GCP, such as the Google Kubernetes Engine API, the Google Container Registry API, and the Google Container File System API, which are necessary for the deployment of the GKE Cluster and the Hugging Face DLC for TEI.",Xe,S,qe,R,Kt="Additionally, to use <code>kubectl</code> with the GKE Cluster credentials, you also need to install the <code>gke-gcloud-auth-plugin</code>, that can be installed with <code>gcloud</code> as follows:",ze,Y,Oe,w,Ke,B,et,D,el="Once everything is set up, you can proceed with the creation of the GKE Cluster and the node pool, which in this case will be a single CPU node as for most of the workloads CPU inference is enough to serve most of the text embeddings models, while it could benefit a lot from GPU serving.",tt,C,lt,X,tl="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.",nt,T,st,q,it,v,ot,z,at,O,ll='For more information please visit <a href="https://cloud.google.com/kubernetes-engine/versioning#specifying_cluster_version" rel="nofollow">https://cloud.google.com/kubernetes-engine/versioning#specifying_cluster_version</a>.',rt,K,nl='<img src="https://raw.githubusercontent.com/huggingface/Google-Cloud-Containers/main/examples/gke/tei-deployment/imgs/gke-cluster.png" alt="GKE Cluster in the GCP Console"/>',ct,ee,sl="Once the GKE Cluster is created, you can get the credentials to access it via <code>kubectl</code> with the following command:",pt,te,ut,le,ft,ne,il='Now you can proceed to the Kubernetes deployment of the Hugging Face DLC for TEI, serving the <a href="https://huggingface.co/Snowflake/snowflake-arctic-embed-m" rel="nofollow"><code>Snowflake/snowflake-arctic-embed-m</code></a> model from the Hugging Face Hub.',dt,U,mt,se,ol="The Hugging Face DLC for TEI will be deployed via <code>kubectl</code>, from the configuration files in either the <code>cpu-config/</code> or the <code>gpu-config/</code> directories depending on whether you want to use the CPU or GPU accelerators, respectively:",gt,ie,al='<li><code>deployment.yaml</code>: contains the deployment details of the pod including the reference to the Hugging Face DLC for TEI setting the <code>MODEL_ID</code> to <a href="https://huggingface.co/Snowflake/snowflake-arctic-embed-m" rel="nofollow"><code>Snowflake/snowflake-arctic-embed-m</code></a>.</li> <li><code>service.yaml</code>: contains the service details of the pod, exposing the port 8080 for the TEI service.</li> <li>(optional) <code>ingress.yaml</code>: contains the ingress details of the pod, exposing the service to the external world so that it can be accessed via the ingress IP.</li>',ht,oe,$t,x,bt,ae,rl='<img src="https://raw.githubusercontent.com/huggingface/Google-Cloud-Containers/main/examples/gke/tei-deployment/imgs/gke-deployment.png" alt="GKE Deployment in the GCP Console"/>',yt,j,Mt,re,wt,ce,cl="Alternatively, you can just wait for the deployment to be ready with the following command:",Ct,pe,Tt,ue,vt,fe,pl="To run the inference over the deployed TEI service, you can either:",Ut,k,de,Ze,ul="Port-forwarding the deployed TEI service to the port 8080, so as to access via <code>localhost</code> with the command:",Qt,me,St,ge,Ve,fl="Accessing the TEI service via the external IP of the ingress, which is the default scenario here since you have defined the ingress configuration in the <code>cpu-config/ingress.yaml</code> or the <code>gpu-config/ingress.yaml</code> file (but it can be skipped in favour of the port-forwarding), that can be retrieved with the following command:",Rt,he,xt,E,jt,$e,dl=`<li><strong>Text Embeddings</strong>: text embedding models expose the endpoint <code>/embed</code> expecting a payload with the key <code>inputs</code> which is either a string or a list of strings to be embedded.</li> <li><strong>Re-rank</strong>: re-ranker models expose the endpoint <code>/rerank</code> expecting a payload with the keys <code>query</code> and <code>texts</code>, where the <code>query</code> is the reference used to rank the similarity against each text in <code>texts</code>.</li> <li><strong>Sequence Classification</strong>: classic sequence classification models expose the endpoint <code>/predict</code> which expects a payload with the key <code>inputs</code> which is either a string or a list of strings to classify.
More information at <a href="https://huggingface.co/docs/text-embeddings-inference/quick_tour" rel="nofollow">https://huggingface.co/docs/text-embeddings-inference/quick_tour</a>.</li>`,kt,be,Et,ye,ml="To send a POST request to the TEI service using <code>cURL</code>, you can run the following command:",Gt,Me,It,we,gl="Or send a POST request to the ingress IP instead:",Jt,Ce,_t,Te,hl='Which produces the following output (truncated for brevity, but original tensor length is 768, which is the embedding dimension of <a href="https://huggingface.co/Snowflake/snowflake-arctic-embed-m" rel="nofollow"><code>Snowflake/snowflake-arctic-embed-m</code></a> i.e. the model you are serving):',Lt,ve,Zt,Ue,Vt,xe,$l="Finally, once you are done using TEI on the GKE Cluster, you can safely delete the GKE Cluster to avoid incurring in unnecessary costs.",Nt,je,Ht,ke,bl="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 <code>e2-small</code> instance.",At,Ee,Pt,Ge,Wt,Ne,Ft;return _=new Le({props:{title:"Deploy Snowflake’s Arctic Embed with TEI DLC on GKE",local:"deploy-snowflakes-arctic-embed-with-tei-dlc-on-gke",headingTag:"h1"}}),V=new Le({props:{title:"Setup / Configuration",local:"setup--configuration",headingTag:"h2"}}),P=new M({props:{code:"ZXhwb3J0JTIwUFJPSkVDVF9JRCUzRHlvdXItcHJvamVjdC1pZCUwQWV4cG9ydCUyMExPQ0FUSU9OJTNEeW91ci1sb2NhdGlvbiUwQWV4cG9ydCUyMENMVVNURVJfTkFNRSUzRHlvdXItY2x1c3Rlci1uYW1l",highlighted:`<span class="hljs-built_in">export</span> PROJECT_ID=your-project-id
<span class="hljs-built_in">export</span> LOCATION=your-location
<span class="hljs-built_in">export</span> CLUSTER_NAME=your-cluster-name`,wrap:!1}}),F=new M({props:{code:"Z2Nsb3VkJTIwYXV0aCUyMGxvZ2luJTBBZ2Nsb3VkJTIwYXV0aCUyMGFwcGxpY2F0aW9uLWRlZmF1bHQlMjBsb2dpbiUyMCUyMCUyMyUyMEZvciUyMGxvY2FsJTIwZGV2ZWxvcG1lbnQlMEFnY2xvdWQlMjBjb25maWclMjBzZXQlMjBwcm9qZWN0JTIwJTI0UFJPSkVDVF9JRA==",highlighted:`gcloud auth login
gcloud auth application-default login  <span class="hljs-comment"># For local development</span>
gcloud config <span class="hljs-built_in">set</span> project <span class="hljs-variable">$PROJECT_ID</span>`,wrap:!1}}),S=new M({props:{code:"Z2Nsb3VkJTIwc2VydmljZXMlMjBlbmFibGUlMjBjb250YWluZXIuZ29vZ2xlYXBpcy5jb20lMEFnY2xvdWQlMjBzZXJ2aWNlcyUyMGVuYWJsZSUyMGNvbnRhaW5lcnJlZ2lzdHJ5Lmdvb2dsZWFwaXMuY29tJTBBZ2Nsb3VkJTIwc2VydmljZXMlMjBlbmFibGUlMjBjb250YWluZXJmaWxlc3lzdGVtLmdvb2dsZWFwaXMuY29t",highlighted:`gcloud services <span class="hljs-built_in">enable</span> container.googleapis.com
gcloud services <span class="hljs-built_in">enable</span> containerregistry.googleapis.com
gcloud services <span class="hljs-built_in">enable</span> containerfilesystem.googleapis.com`,wrap:!1}}),Y=new M({props:{code:"Z2Nsb3VkJTIwY29tcG9uZW50cyUyMGluc3RhbGwlMjBna2UtZ2Nsb3VkLWF1dGgtcGx1Z2lu",highlighted:"gcloud components install gke-gcloud-auth-plugin",wrap:!1}}),w=new I({props:{$$slots:{default:[Il]},$$scope:{ctx:y}}}),B=new Le({props:{title:"Create GKE Cluster",local:"create-gke-cluster",headingTag:"h2"}}),C=new I({props:{$$slots:{default:[Jl]},$$scope:{ctx:y}}}),T=new I({props:{$$slots:{default:[_l]},$$scope:{ctx:y}}}),q=new M({props:{code:"Z2Nsb3VkJTIwY29udGFpbmVyJTIwY2x1c3RlcnMlMjBjcmVhdGUtYXV0byUyMCUyNENMVVNURVJfTkFNRSUyMCU1QyUwQSUyMCUyMCUyMCUyMC0tcHJvamVjdCUzRCUyNFBST0pFQ1RfSUQlMjAlNUMlMEElMjAlMjAlMjAlMjAtLWxvY2F0aW9uJTNEJTI0TE9DQVRJT04lMjAlNUMlMEElMjAlMjAlMjAlMjAtLXJlbGVhc2UtY2hhbm5lbCUzRHN0YWJsZSUyMCU1QyUwQSUyMCUyMCUyMCUyMC0tY2x1c3Rlci12ZXJzaW9uJTNEMS4yOCUyMCU1QyUwQSUyMCUyMCUyMCUyMC0tbm8tYXV0b3Byb3Zpc2lvbmluZy1lbmFibGUtaW5zZWN1cmUta3ViZWxldC1yZWFkb25seS1wb3J0",highlighted:`gcloud container clusters create-auto <span class="hljs-variable">$CLUSTER_NAME</span> \\
    --project=<span class="hljs-variable">$PROJECT_ID</span> \\
    --location=<span class="hljs-variable">$LOCATION</span> \\
    --release-channel=stable \\
    --cluster-version=1.28 \\
    --no-autoprovisioning-enable-insecure-kubelet-readonly-port`,wrap:!1}}),v=new I({props:{$$slots:{default:[Ll]},$$scope:{ctx:y}}}),z=new M({props:{code:"Z2Nsb3VkJTIwY29udGFpbmVyJTIwZ2V0LXNlcnZlci1jb25maWclMjAlNUMlMEEtLWZsYXR0ZW4lM0QlMjJjaGFubmVscyUyMiUyMCU1QyUwQS0tZmlsdGVyJTNEJTIyY2hhbm5lbHMuY2hhbm5lbCUzRFNUQUJMRSUyMiUyMCU1QyUwQS0tZm9ybWF0JTNEJTIyeWFtbChjaGFubmVscy5jaGFubmVsJTJDY2hhbm5lbHMuZGVmYXVsdFZlcnNpb24pJTIyJTIwJTVDJTBBLS1sb2NhdGlvbiUzRCUyNExPQ0FUSU9O",highlighted:`gcloud container get-server-config \\
--flatten=<span class="hljs-string">&quot;channels&quot;</span> \\
--filter=<span class="hljs-string">&quot;channels.channel=STABLE&quot;</span> \\
--format=<span class="hljs-string">&quot;yaml(channels.channel,channels.defaultVersion)&quot;</span> \\
--location=<span class="hljs-variable">$LOCATION</span>`,wrap:!1}}),te=new M({props:{code:"Z2Nsb3VkJTIwY29udGFpbmVyJTIwY2x1c3RlcnMlMjBnZXQtY3JlZGVudGlhbHMlMjAlMjRDTFVTVEVSX05BTUUlMjAtLWxvY2F0aW9uJTNEJTI0TE9DQVRJT04=",highlighted:'gcloud container clusters get-credentials <span class="hljs-variable">$CLUSTER_NAME</span> --location=<span class="hljs-variable">$LOCATION</span>',wrap:!1}}),le=new Le({props:{title:"Deploy TEI",local:"deploy-tei",headingTag:"h2"}}),U=new I({props:{$$slots:{default:[Zl]},$$scope:{ctx:y}}}),oe=new M({props:{code:"a3ViZWN0bCUyMGFwcGx5JTIwLWYlMjBjcHUtY29uZmlnJTJG",highlighted:"kubectl apply -f cpu-config/",wrap:!1}}),x=new I({props:{$$slots:{default:[Vl]},$$scope:{ctx:y}}}),j=new I({props:{$$slots:{default:[Nl]},$$scope:{ctx:y}}}),re=new M({props:{code:"a3ViZWN0bCUyMGdldCUyMHBvZHM=",highlighted:"kubectl get pods",wrap:!1}}),pe=new M({props:{code:"a3ViZWN0bCUyMHdhaXQlMjAtLWZvciUzRGNvbmRpdGlvbiUzREF2YWlsYWJsZSUyMC0tdGltZW91dCUzRDcwMHMlMjBkZXBsb3ltZW50JTJGdGVpLWRlcGxveW1lbnQ=",highlighted:'kubectl <span class="hljs-built_in">wait</span> --<span class="hljs-keyword">for</span>=condition=Available --<span class="hljs-built_in">timeout</span>=700s deployment/tei-deployment',wrap:!1}}),ue=new Le({props:{title:"Inference with TEI",local:"inference-with-tei",headingTag:"h2"}}),me=new M({props:{code:"a3ViZWN0bCUyMHBvcnQtZm9yd2FyZCUyMHNlcnZpY2UlMkZ0ZWktc2VydmljZSUyMDgwODAlM0E4MDgw",highlighted:"kubectl port-forward service/tei-service 8080:8080",wrap:!1}}),he=new M({props:{code:"a3ViZWN0bCUyMGdldCUyMGluZ3Jlc3MlMjB0ZWktaW5ncmVzcyUyMC1vJTIwanNvbnBhdGglM0QnJTdCLnN0YXR1cy5sb2FkQmFsYW5jZXIuaW5ncmVzcyU1QjAlNUQuaXAlN0Qn",highlighted:'kubectl get ingress tei-ingress -o jsonpath=<span class="hljs-string">&#x27;{.status.loadBalancer.ingress[0].ip}&#x27;</span>',wrap:!1}}),E=new I({props:{$$slots:{default:[Hl]},$$scope:{ctx:y}}}),be=new Le({props:{title:"Via cURL",local:"via-curl",headingTag:"h3"}}),Me=new M({props:{code:"Y3VybCUyMGh0dHAlM0ElMkYlMkZsb2NhbGhvc3QlM0E4MDgwJTJGZW1iZWQlMjAlNUMlMEElMjAlMjAlMjAlMjAtWCUyMFBPU1QlMjAlNUMlMEElMjAlMjAlMjAlMjAtZCUyMCclN0IlMjJpbnB1dHMlMjIlM0ElMjJXaGF0JTIwaXMlMjBEZWVwJTIwTGVhcm5pbmclM0YlMjIlN0QnJTIwJTVDJTBBJTIwJTIwJTIwJTIwLUglMjAnQ29udGVudC1UeXBlJTNBJTIwYXBwbGljYXRpb24lMkZqc29uJw==",highlighted:`curl http://localhost:8080/embed \\
    -X POST \\
    -d <span class="hljs-string">&#x27;{&quot;inputs&quot;:&quot;What is Deep Learning?&quot;}&#x27;</span> \\
    -H <span class="hljs-string">&#x27;Content-Type: application/json&#x27;</span>`,wrap:!1}}),Ce=new M({props:{code:"Y3VybCUyMGh0dHAlM0ElMkYlMkYlMjQoa3ViZWN0bCUyMGdldCUyMGluZ3Jlc3MlMjB0ZWktaW5ncmVzcyUyMC1vJTIwanNvbnBhdGglM0QnJTdCLnN0YXR1cy5sb2FkQmFsYW5jZXIuaW5ncmVzcyU1QjAlNUQuaXAlN0QnKSUzQTgwODAlMkZlbWJlZCUyMCU1QyUwQSUyMCUyMCUyMCUyMC1YJTIwUE9TVCUyMCU1QyUwQSUyMCUyMCUyMCUyMC1kJTIwJyU3QiUyMmlucHV0cyUyMiUzQSUyMldoYXQlMjBpcyUyMERlZXAlMjBMZWFybmluZyUzRiUyMiU3RCclMjAlNUMlMEElMjAlMjAlMjAlMjAtSCUyMCdDb250ZW50LVR5cGUlM0ElMjBhcHBsaWNhdGlvbiUyRmpzb24n",highlighted:`curl http://$(kubectl get ingress tei-ingress -o jsonpath=<span class="hljs-string">&#x27;{.status.loadBalancer.ingress[0].ip}&#x27;</span>):8080/embed \\
    -X POST \\
    -d <span class="hljs-string">&#x27;{&quot;inputs&quot;:&quot;What is Deep Learning?&quot;}&#x27;</span> \\
    -H <span class="hljs-string">&#x27;Content-Type: application/json&#x27;</span>`,wrap:!1}}),ve=new M({props:{code:"JTVCJTVCLTAuMDE0ODMwOTglMkMwLjAxMDg0NjM1OSUyQy0wLjAyNDY3OTIzNiUyQzAuMDEyNTA3NjI4JTJDMC4wMzQyMzE1NTUlMkMuLi4lNUQlNUQ=",highlighted:'[[-<span class="hljs-number">0.01483098</span>,<span class="hljs-number">0</span>.<span class="hljs-number">010846359</span>,-<span class="hljs-number">0.024679236</span>,<span class="hljs-number">0</span>.<span class="hljs-number">012507628</span>,<span class="hljs-number">0.034231555</span>,...]]',wrap:!1}}),Ue=new Le({props:{title:"Delete GKE Cluster",local:"delete-gke-cluster",headingTag:"h2"}}),je=new M({props:{code:"Z2Nsb3VkJTIwY29udGFpbmVyJTIwY2x1c3RlcnMlMjBkZWxldGUlMjAlMjRDTFVTVEVSX05BTUUlMjAtLWxvY2F0aW9uJTNEJTI0TE9DQVRJT04=",highlighted:'gcloud container clusters delete <span class="hljs-variable">$CLUSTER_NAME</span> --location=<span class="hljs-variable">$LOCATION</span>',wrap:!1}}),Ee=new M({props:{code:"a3ViZWN0bCUyMHNjYWxlJTIwLS1yZXBsaWNhcyUzRDAlMjBkZXBsb3ltZW50JTJGdGVpLWRlcGxveW1lbnQ=",highlighted:"kubectl scale --replicas=0 deployment/tei-deployment",wrap:!1}}),Ge=new Gl({props:{source:"https://github.com/huggingface/Google-Cloud-Containers/blob/main/docs/source/examples/gke-tei-deployment.mdx"}}),{c(){s=r("meta"),$=i(),a=r("p"),b=i(),u(_.$$.fragment),Ae=i(),L=r("p"),L.textContent=Yt,Pe=i(),Z=r("p"),Z.textContent=Bt,We=i(),u(V.$$.fragment),Fe=i(),N=r("p"),N.innerHTML=Dt,Qe=i(),H=r("ul"),H.innerHTML=Xt,Se=i(),A=r("p"),A.textContent=qt,Re=i(),u(P.$$.fragment),Ye=i(),W=r("p"),W.textContent=zt,Be=i(),u(F.$$.fragment),De=i(),Q=r("p"),Q.textContent=Ot,Xe=i(),u(S.$$.fragment),qe=i(),R=r("p"),R.innerHTML=Kt,ze=i(),u(Y.$$.fragment),Oe=i(),u(w.$$.fragment),Ke=i(),u(B.$$.fragment),et=i(),D=r("p"),D.textContent=el,tt=i(),u(C.$$.fragment),lt=i(),X=r("p"),X.textContent=tl,nt=i(),u(T.$$.fragment),st=i(),u(q.$$.fragment),it=i(),u(v.$$.fragment),ot=i(),u(z.$$.fragment),at=i(),O=r("p"),O.innerHTML=ll,rt=i(),K=r("p"),K.innerHTML=nl,ct=i(),ee=r("p"),ee.innerHTML=sl,pt=i(),u(te.$$.fragment),ut=i(),u(le.$$.fragment),ft=i(),ne=r("p"),ne.innerHTML=il,dt=i(),u(U.$$.fragment),mt=i(),se=r("p"),se.innerHTML=ol,gt=i(),ie=r("ul"),ie.innerHTML=al,ht=i(),u(oe.$$.fragment),$t=i(),u(x.$$.fragment),bt=i(),ae=r("p"),ae.innerHTML=rl,yt=i(),u(j.$$.fragment),Mt=i(),u(re.$$.fragment),wt=i(),ce=r("p"),ce.textContent=cl,Ct=i(),u(pe.$$.fragment),Tt=i(),u(ue.$$.fragment),vt=i(),fe=r("p"),fe.textContent=pl,Ut=i(),k=r("ul"),de=r("li"),Ze=r("p"),Ze.innerHTML=ul,Qt=i(),u(me.$$.fragment),St=i(),ge=r("li"),Ve=r("p"),Ve.innerHTML=fl,Rt=i(),u(he.$$.fragment),xt=i(),u(E.$$.fragment),jt=i(),$e=r("ul"),$e.innerHTML=dl,kt=i(),u(be.$$.fragment),Et=i(),ye=r("p"),ye.innerHTML=ml,Gt=i(),u(Me.$$.fragment),It=i(),we=r("p"),we.textContent=gl,Jt=i(),u(Ce.$$.fragment),_t=i(),Te=r("p"),Te.innerHTML=hl,Lt=i(),u(ve.$$.fragment),Zt=i(),u(Ue.$$.fragment),Vt=i(),xe=r("p"),xe.textContent=$l,Nt=i(),u(je.$$.fragment),Ht=i(),ke=r("p"),ke.innerHTML=bl,At=i(),u(Ee.$$.fragment),Pt=i(),u(Ge.$$.fragment),Wt=i(),Ne=r("p"),this.h()},l(e){const t=El("svelte-u9bgzb",document.head);s=c(t,"META",{name:!0,content:!0}),t.forEach(l),$=o(e),a=c(e,"P",{}),He(a).forEach(l),b=o(e),f(_.$$.fragment,e),Ae=o(e),L=c(e,"P",{"data-svelte-h":!0}),p(L)!=="svelte-tipapj"&&(L.textContent=Yt),Pe=o(e),Z=c(e,"P",{"data-svelte-h":!0}),p(Z)!=="svelte-qfbuq0"&&(Z.textContent=Bt),We=o(e),f(V.$$.fragment,e),Fe=o(e),N=c(e,"P",{"data-svelte-h":!0}),p(N)!=="svelte-1qnmo38"&&(N.innerHTML=Dt),Qe=o(e),H=c(e,"UL",{"data-svelte-h":!0}),p(H)!=="svelte-a8ymuw"&&(H.innerHTML=Xt),Se=o(e),A=c(e,"P",{"data-svelte-h":!0}),p(A)!=="svelte-1i8n6rq"&&(A.textContent=qt),Re=o(e),f(P.$$.fragment,e),Ye=o(e),W=c(e,"P",{"data-svelte-h":!0}),p(W)!=="svelte-1g39t6p"&&(W.textContent=zt),Be=o(e),f(F.$$.fragment,e),De=o(e),Q=c(e,"P",{"data-svelte-h":!0}),p(Q)!=="svelte-ly6a6j"&&(Q.textContent=Ot),Xe=o(e),f(S.$$.fragment,e),qe=o(e),R=c(e,"P",{"data-svelte-h":!0}),p(R)!=="svelte-wjaltb"&&(R.innerHTML=Kt),ze=o(e),f(Y.$$.fragment,e),Oe=o(e),f(w.$$.fragment,e),Ke=o(e),f(B.$$.fragment,e),et=o(e),D=c(e,"P",{"data-svelte-h":!0}),p(D)!=="svelte-1202via"&&(D.textContent=el),tt=o(e),f(C.$$.fragment,e),lt=o(e),X=c(e,"P",{"data-svelte-h":!0}),p(X)!=="svelte-l7bdz6"&&(X.textContent=tl),nt=o(e),f(T.$$.fragment,e),st=o(e),f(q.$$.fragment,e),it=o(e),f(v.$$.fragment,e),ot=o(e),f(z.$$.fragment,e),at=o(e),O=c(e,"P",{"data-svelte-h":!0}),p(O)!=="svelte-1y4gd42"&&(O.innerHTML=ll),rt=o(e),K=c(e,"P",{"data-svelte-h":!0}),p(K)!=="svelte-1owrs4z"&&(K.innerHTML=nl),ct=o(e),ee=c(e,"P",{"data-svelte-h":!0}),p(ee)!=="svelte-f29e6h"&&(ee.innerHTML=sl),pt=o(e),f(te.$$.fragment,e),ut=o(e),f(le.$$.fragment,e),ft=o(e),ne=c(e,"P",{"data-svelte-h":!0}),p(ne)!=="svelte-6opf9o"&&(ne.innerHTML=il),dt=o(e),f(U.$$.fragment,e),mt=o(e),se=c(e,"P",{"data-svelte-h":!0}),p(se)!=="svelte-wsnwcj"&&(se.innerHTML=ol),gt=o(e),ie=c(e,"UL",{"data-svelte-h":!0}),p(ie)!=="svelte-jkdrgr"&&(ie.innerHTML=al),ht=o(e),f(oe.$$.fragment,e),$t=o(e),f(x.$$.fragment,e),bt=o(e),ae=c(e,"P",{"data-svelte-h":!0}),p(ae)!=="svelte-xnhm25"&&(ae.innerHTML=rl),yt=o(e),f(j.$$.fragment,e),Mt=o(e),f(re.$$.fragment,e),wt=o(e),ce=c(e,"P",{"data-svelte-h":!0}),p(ce)!=="svelte-16v3quh"&&(ce.textContent=cl),Ct=o(e),f(pe.$$.fragment,e),Tt=o(e),f(ue.$$.fragment,e),vt=o(e),fe=c(e,"P",{"data-svelte-h":!0}),p(fe)!=="svelte-1is499s"&&(fe.textContent=pl),Ut=o(e),k=c(e,"UL",{});var Ie=He(k);de=c(Ie,"LI",{});var Je=He(de);Ze=c(Je,"P",{"data-svelte-h":!0}),p(Ze)!=="svelte-1pbh49h"&&(Ze.innerHTML=ul),Qt=o(Je),f(me.$$.fragment,Je),Je.forEach(l),St=o(Ie),ge=c(Ie,"LI",{});var _e=He(ge);Ve=c(_e,"P",{"data-svelte-h":!0}),p(Ve)!=="svelte-kfuy0z"&&(Ve.innerHTML=fl),Rt=o(_e),f(he.$$.fragment,_e),_e.forEach(l),Ie.forEach(l),xt=o(e),f(E.$$.fragment,e),jt=o(e),$e=c(e,"UL",{"data-svelte-h":!0}),p($e)!=="svelte-1dffo21"&&($e.innerHTML=dl),kt=o(e),f(be.$$.fragment,e),Et=o(e),ye=c(e,"P",{"data-svelte-h":!0}),p(ye)!=="svelte-189d42i"&&(ye.innerHTML=ml),Gt=o(e),f(Me.$$.fragment,e),It=o(e),we=c(e,"P",{"data-svelte-h":!0}),p(we)!=="svelte-1u6gmmt"&&(we.textContent=gl),Jt=o(e),f(Ce.$$.fragment,e),_t=o(e),Te=c(e,"P",{"data-svelte-h":!0}),p(Te)!=="svelte-2igbfl"&&(Te.innerHTML=hl),Lt=o(e),f(ve.$$.fragment,e),Zt=o(e),f(Ue.$$.fragment,e),Vt=o(e),xe=c(e,"P",{"data-svelte-h":!0}),p(xe)!=="svelte-1teoy12"&&(xe.textContent=$l),Nt=o(e),f(je.$$.fragment,e),Ht=o(e),ke=c(e,"P",{"data-svelte-h":!0}),p(ke)!=="svelte-qo8r7n"&&(ke.innerHTML=bl),At=o(e),f(Ee.$$.fragment,e),Pt=o(e),f(Ge.$$.fragment,e),Wt=o(e),Ne=c(e,"P",{}),He(Ne).forEach(l),this.h()},h(){vl(s,"name","hf:doc:metadata"),vl(s,"content",Pl)},m(e,t){G(document.head,s),n(e,$,t),n(e,a,t),n(e,b,t),d(_,e,t),n(e,Ae,t),n(e,L,t),n(e,Pe,t),n(e,Z,t),n(e,We,t),d(V,e,t),n(e,Fe,t),n(e,N,t),n(e,Qe,t),n(e,H,t),n(e,Se,t),n(e,A,t),n(e,Re,t),d(P,e,t),n(e,Ye,t),n(e,W,t),n(e,Be,t),d(F,e,t),n(e,De,t),n(e,Q,t),n(e,Xe,t),d(S,e,t),n(e,qe,t),n(e,R,t),n(e,ze,t),d(Y,e,t),n(e,Oe,t),d(w,e,t),n(e,Ke,t),d(B,e,t),n(e,et,t),n(e,D,t),n(e,tt,t),d(C,e,t),n(e,lt,t),n(e,X,t),n(e,nt,t),d(T,e,t),n(e,st,t),d(q,e,t),n(e,it,t),d(v,e,t),n(e,ot,t),d(z,e,t),n(e,at,t),n(e,O,t),n(e,rt,t),n(e,K,t),n(e,ct,t),n(e,ee,t),n(e,pt,t),d(te,e,t),n(e,ut,t),d(le,e,t),n(e,ft,t),n(e,ne,t),n(e,dt,t),d(U,e,t),n(e,mt,t),n(e,se,t),n(e,gt,t),n(e,ie,t),n(e,ht,t),d(oe,e,t),n(e,$t,t),d(x,e,t),n(e,bt,t),n(e,ae,t),n(e,yt,t),d(j,e,t),n(e,Mt,t),d(re,e,t),n(e,wt,t),n(e,ce,t),n(e,Ct,t),d(pe,e,t),n(e,Tt,t),d(ue,e,t),n(e,vt,t),n(e,fe,t),n(e,Ut,t),n(e,k,t),G(k,de),G(de,Ze),G(de,Qt),d(me,de,null),G(k,St),G(k,ge),G(ge,Ve),G(ge,Rt),d(he,ge,null),n(e,xt,t),d(E,e,t),n(e,jt,t),n(e,$e,t),n(e,kt,t),d(be,e,t),n(e,Et,t),n(e,ye,t),n(e,Gt,t),d(Me,e,t),n(e,It,t),n(e,we,t),n(e,Jt,t),d(Ce,e,t),n(e,_t,t),n(e,Te,t),n(e,Lt,t),d(ve,e,t),n(e,Zt,t),d(Ue,e,t),n(e,Vt,t),n(e,xe,t),n(e,Nt,t),d(je,e,t),n(e,Ht,t),n(e,ke,t),n(e,At,t),d(Ee,e,t),n(e,Pt,t),d(Ge,e,t),n(e,Wt,t),n(e,Ne,t),Ft=!0},p(e,[t]){const Ie={};t&2&&(Ie.$$scope={dirty:t,ctx:e}),w.$set(Ie);const Je={};t&2&&(Je.$$scope={dirty:t,ctx:e}),C.$set(Je);const _e={};t&2&&(_e.$$scope={dirty:t,ctx:e}),T.$set(_e);const yl={};t&2&&(yl.$$scope={dirty:t,ctx:e}),v.$set(yl);const Ml={};t&2&&(Ml.$$scope={dirty:t,ctx:e}),U.$set(Ml);const wl={};t&2&&(wl.$$scope={dirty:t,ctx:e}),x.$set(wl);const Cl={};t&2&&(Cl.$$scope={dirty:t,ctx:e}),j.$set(Cl);const Tl={};t&2&&(Tl.$$scope={dirty:t,ctx:e}),E.$set(Tl)},i(e){Ft||(m(_.$$.fragment,e),m(V.$$.fragment,e),m(P.$$.fragment,e),m(F.$$.fragment,e),m(S.$$.fragment,e),m(Y.$$.fragment,e),m(w.$$.fragment,e),m(B.$$.fragment,e),m(C.$$.fragment,e),m(T.$$.fragment,e),m(q.$$.fragment,e),m(v.$$.fragment,e),m(z.$$.fragment,e),m(te.$$.fragment,e),m(le.$$.fragment,e),m(U.$$.fragment,e),m(oe.$$.fragment,e),m(x.$$.fragment,e),m(j.$$.fragment,e),m(re.$$.fragment,e),m(pe.$$.fragment,e),m(ue.$$.fragment,e),m(me.$$.fragment,e),m(he.$$.fragment,e),m(E.$$.fragment,e),m(be.$$.fragment,e),m(Me.$$.fragment,e),m(Ce.$$.fragment,e),m(ve.$$.fragment,e),m(Ue.$$.fragment,e),m(je.$$.fragment,e),m(Ee.$$.fragment,e),m(Ge.$$.fragment,e),Ft=!0)},o(e){g(_.$$.fragment,e),g(V.$$.fragment,e),g(P.$$.fragment,e),g(F.$$.fragment,e),g(S.$$.fragment,e),g(Y.$$.fragment,e),g(w.$$.fragment,e),g(B.$$.fragment,e),g(C.$$.fragment,e),g(T.$$.fragment,e),g(q.$$.fragment,e),g(v.$$.fragment,e),g(z.$$.fragment,e),g(te.$$.fragment,e),g(le.$$.fragment,e),g(U.$$.fragment,e),g(oe.$$.fragment,e),g(x.$$.fragment,e),g(j.$$.fragment,e),g(re.$$.fragment,e),g(pe.$$.fragment,e),g(ue.$$.fragment,e),g(me.$$.fragment,e),g(he.$$.fragment,e),g(E.$$.fragment,e),g(be.$$.fragment,e),g(Me.$$.fragment,e),g(Ce.$$.fragment,e),g(ve.$$.fragment,e),g(Ue.$$.fragment,e),g(je.$$.fragment,e),g(Ee.$$.fragment,e),g(Ge.$$.fragment,e),Ft=!1},d(e){e&&(l($),l(a),l(b),l(Ae),l(L),l(Pe),l(Z),l(We),l(Fe),l(N),l(Qe),l(H),l(Se),l(A),l(Re),l(Ye),l(W),l(Be),l(De),l(Q),l(Xe),l(qe),l(R),l(ze),l(Oe),l(Ke),l(et),l(D),l(tt),l(lt),l(X),l(nt),l(st),l(it),l(ot),l(at),l(O),l(rt),l(K),l(ct),l(ee),l(pt),l(ut),l(ft),l(ne),l(dt),l(mt),l(se),l(gt),l(ie),l(ht),l($t),l(bt),l(ae),l(yt),l(Mt),l(wt),l(ce),l(Ct),l(Tt),l(vt),l(fe),l(Ut),l(k),l(xt),l(jt),l($e),l(kt),l(Et),l(ye),l(Gt),l(It),l(we),l(Jt),l(_t),l(Te),l(Lt),l(Zt),l(Vt),l(xe),l(Nt),l(Ht),l(ke),l(At),l(Pt),l(Wt),l(Ne)),l(s),h(_,e),h(V,e),h(P,e),h(F,e),h(S,e),h(Y,e),h(w,e),h(B,e),h(C,e),h(T,e),h(q,e),h(v,e),h(z,e),h(te,e),h(le,e),h(U,e),h(oe,e),h(x,e),h(j,e),h(re,e),h(pe,e),h(ue,e),h(me),h(he),h(E,e),h(be,e),h(Me,e),h(Ce,e),h(ve,e),h(Ue,e),h(je,e),h(Ee,e),h(Ge,e)}}}const Pl='{"title":"Deploy Snowflake’s Arctic Embed with TEI DLC on GKE","local":"deploy-snowflakes-arctic-embed-with-tei-dlc-on-gke","sections":[{"title":"Setup / Configuration","local":"setup--configuration","sections":[],"depth":2},{"title":"Create GKE Cluster","local":"create-gke-cluster","sections":[],"depth":2},{"title":"Deploy TEI","local":"deploy-tei","sections":[],"depth":2},{"title":"Inference with TEI","local":"inference-with-tei","sections":[{"title":"Via cURL","local":"via-curl","sections":[],"depth":3}],"depth":2},{"title":"Delete GKE Cluster","local":"delete-gke-cluster","sections":[],"depth":2}],"depth":1}';function Wl(y){return xl(()=>{new URLSearchParams(window.location.search).get("fw")}),[]}class Bl extends jl{constructor(s){super(),kl(this,s,Wl,Al,Ul,{})}}export{Bl as component};