diff --git a/third_party/Dell/model-deployment/Qwen3-8b/deployment.md b/third_party/Dell/model-deployment/Qwen3-8b/deployment.md new file mode 100644 index 00000000..f989426f --- /dev/null +++ b/third_party/Dell/model-deployment/Qwen3-8b/deployment.md @@ -0,0 +1,101 @@ +## Step 1: Prerequisites to Deploy Qwen3-8b Model on Xeon with Keycloak + +Ensure the Enterprise Inference stack with Keycloak is already deployed before proceeding. + +Edit `core/scripts/generate-token.sh` and set your values before sourcing it: + +| Variable | Description | +| ------------------------- | ------------------------------------------------------------------------ | +| `BASE_URL` | Hostname of your cluster (e.g. `api.example.com`), without `https://` | +| `KEYCLOAK_ADMIN_USERNAME` | Keycloak admin username | +| `KEYCLOAK_PASSWORD` | Keycloak admin password | +| `KEYCLOAK_CLIENT_ID` | Keycloak client ID configured during EI deployment | + +Then run: + +```bash +export HUGGING_FACE_HUB_TOKEN="your_token_here" + +cd ~/Enterprise-Inference +source core/scripts/generate-token.sh +``` + +This exports: `BASE_URL`, `KEYCLOAK_CLIENT_ID`, `KEYCLOAK_CLIENT_SECRET`, and `TOKEN`. + +## Step 2: Deploy Qwen3-8b Model + +```bash +helm install qwen3-8b-cpu ./core/helm-charts/vllm \ + --values ./core/helm-charts/vllm/xeon-values.yaml \ + --set LLM_MODEL_ID="Qwen/Qwen3-8B" \ + --set global.HUGGINGFACEHUB_API_TOKEN="$HUGGING_FACE_HUB_TOKEN" \ + --set ingress.enabled=true \ + --set ingress.secretname="${BASE_URL}" \ + --set ingress.host="${BASE_URL}" \ + --set oidc.client_id="$KEYCLOAK_CLIENT_ID" \ + --set oidc.client_secret="$KEYCLOAK_CLIENT_SECRET" \ + --set apisix.enabled=true \ + --set tensor_parallel_size="1" \ + --set pipeline_parallel_size="1" +``` + +## Step 3: Verify the Deployment + +```bash +kubectl get pods +kubectl get apisixroutes +``` + +Expected Output: + +``` +NAME READY STATUS RESTARTS +keycloak-0 1/1 Running 0 +keycloak-postgresql-0 1/1 Running 0 +qwen3-8b-cpu-vllm-- 1/1 Running 0 +``` + +> Note: The pod name suffix `-` is auto-generated by Kubernetes and will differ on each deployment. Ensure all pods show `1/1 Running`. + +``` +NAME HOSTS +qwen3-8b-cpu-vllm-apisixroute api.example.com +``` + +## Step 4: Test the Deployed Model + +```bash +curl -k https://${BASE_URL}/Qwen3-8B-vllmcpu/v1/completions \ + -X POST \ + -H "Content-Type: application/json" \ + -H "Authorization: Bearer $TOKEN" \ + -d '{ + "model": "Qwen/Qwen3-8B", + "prompt": "What is Deep Learning?", + "max_tokens": 25, + "temperature": 0 + }' +``` + +If successful, the model will return a completion response. + +## To undeploy the model + +```bash +helm uninstall qwen3-8b-cpu +``` + +## Parameters + +| Parameter | Description | +| ----------------------------------------------- | ------------------------------------------------------------------------------------------------- | +| `--set LLM_MODEL_ID="Qwen/Qwen3-8B"` | Defines the target model from **Hugging Face** to deploy. | +| `--set global.HUGGINGFACEHUB_API_TOKEN="..."` | Authenticates access to gated or private Hugging Face models. Replace with your own secure token. | +| `--set ingress.enabled=true` | Enables Kubernetes **Ingress** to expose the model service externally. | +| `--set ingress.host="${BASE_URL}"` | Public hostname or FQDN for the inference endpoint (maps to your Ingress controller IP). | +| `--set ingress.secretname="${BASE_URL}"` | Kubernetes **TLS Secret** used for HTTPS termination at the ingress layer. | +| `--set oidc.client_id="..."` | Keycloak OIDC client ID used for token-based authentication. | +| `--set oidc.client_secret="..."` | Keycloak OIDC client secret corresponding to the client ID. | +| `--set apisix.enabled=true` | Enables **APISIX** as the API gateway for routing and authentication. | +| `--set tensor_parallel_size="1"` | Number of tensor parallel workers. Set to the number of available CPUs/GPUs per node. | +| `--set pipeline_parallel_size="1"` | Number of pipeline parallel stages. Typically `1` for single-node deployments. | diff --git a/third_party/Dell/model-deployment/Qwen3-8b/model-card.md b/third_party/Dell/model-deployment/Qwen3-8b/model-card.md new file mode 100644 index 00000000..76d15973 --- /dev/null +++ b/third_party/Dell/model-deployment/Qwen3-8b/model-card.md @@ -0,0 +1,62 @@ +# Qwen3-8B + +This model uses Qwen3-8B, a large-scale open-weight language model developed by the Qwen Team at Alibaba Cloud. It is designed for high-quality natural language understanding, reasoning, instruction following, and code intelligence across a broad range of enterprise and research workloads. Qwen3-8B represents the next-generation evolution of the Qwen model family, with improved reasoning depth, instruction alignment, and multilingual capabilities. + +For full details including model specifications, licensing, intended use, safety guidance, and example prompts, please visit the official Hugging Face page: **Official Hugging Face Page** + +https://huggingface.co/Qwen/Qwen3-8B + +This model provides inference services only; weights are hosted by Hugging Face under the Qwen License. + +Ensure compliance with the Qwen License terms before using this model. + +### Model Attribution + +**Developer:** Alibaba Cloud / Qwen Team + +**purpose:** General-purpose instruction-tuned reasoning and language model + +**Sizes/Variants:** 8B parameters + +**Modalities:** Text → Natural Language + Code + +**Parameter Size:** 8 Billion + +**Max Context:** Up to ~128K tokens (depending on backend integration) + +**License:** Qwen License (commercial use permitted with conditions) + +### Usage Notice + +**By using this model, you agree that:** + +- Inputs and outputs are processed by the Qwen3-8B model under the Qwen License. +- You are responsible for validating outputs before production use. +- This model should not be used for generating malicious, deceptive, or unsafe content. +- Commercial usage must comply with all Qwen license obligations and regional legal requirements.. + +### Intended Applications + +- Enterprise chatbots and virtual assistants +- Retrieval-Augmented Generation (RAG) systems +- Agentic AI workflows and task automation +- Code generation, debugging, and refactoring +- API reasoning and architecture guidance +- Multilingual document analysis and summarization +- Knowledge base and search augmentation systems + +### Limitations + +- Higher compute and memory requirements than sub-3B models +- May hallucinate in open-ended or low-context prompts +- Not suitable for unsupervised safety-critical decision systems +- Long-context performance depends on serving backend configuration +- Requires responsible deployment with output validation + +### References + +Qwen Project Official Repository - https://github.com/QwenLM + +Hugging Face Model Page — https://huggingface.co/Qwen/Qwen3-8B + +Qwen License - https://github.com/QwenLM/Qwen/blob/main/LICENSE diff --git a/third_party/Dell/model-deployment/README.md b/third_party/Dell/model-deployment/README.md deleted file mode 100644 index 43d98118..00000000 --- a/third_party/Dell/model-deployment/README.md +++ /dev/null @@ -1 +0,0 @@ -# PLACEHOLDER \ No newline at end of file