This demo shows how to deploy the Mistral 7B model on 2xT4 GPUs.
- Deploy the Granite 7B model on 2xT4 GPUs:
kubectl apply -k overlays/granite-7B-
Remember to add your HUGGING_FACE_HUB_TOKEN into the Environment Variables to be able to download the model from the Hugging Face Hub.
-
Check that the LLM is running properly:
kubectl get pod -n multi-gpu-poc
NAME READY STATUS RESTARTS AGE
llm1-f687846b9-68bvq 1/1 Running 0 2m1s- Check the logs of the Pod LLM:
kubectl logs -n multi-gpu-poc -l app=llm1The output should be similar to:
Avg prompt throughput: 0.0 tokens/s, Avg generation throughput: 0.0 tokens/s, Running: 0 reqs, Swapped: 0 reqs, Pending: 0 reqs, GPU KV cache usage: 0.0%, CPU KV cache usage: 0.0%- Check the NVIDIA GPU consumption:
POD_NAME=$(kubectl get pod -n nvidia-gpu-operator -l app=nvidia-device-plugin-daemonset -o jsonpath="{.items[0].metadata.name}")
kubectl exec -n nvidia-gpu-operator $POD_NAME -- nvidia-smi- The output should be similar to:
+---------------------------------------------------------------------------------------+
| NVIDIA-SMI 535.104.05 Driver Version: 535.104.05 CUDA Version: 12.2 |
|-----------------------------------------+----------------------+----------------------+
| GPU Name Persistence-M | Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|=========================================+======================+======================|
| 0 Tesla T4 On | 00000000:00:1B.0 Off | 0 |
| N/A 30C P8 14W / 70W | 0MiB / 15360MiB | 0% Default |
| | | N/A |
+-----------------------------------------+----------------------+----------------------+
| 1 Tesla T4 On | 00000000:00:1C.0 Off | 0 |
| N/A 30C P8 14W / 70W | 0MiB / 15360MiB | 0% Default |
| | | N/A |
+-----------------------------------------+----------------------+----------------------+
| 2 Tesla T4 On | 00000000:00:1D.0 Off | 0 |
| N/A 39C P0 27W / 70W | 14129MiB / 15360MiB | 0% Default |
| | | N/A |
+-----------------------------------------+----------------------+----------------------+
| 3 Tesla T4 On | 00000000:00:1E.0 Off | 0 |
| N/A 38C P0 26W / 70W | 14101MiB / 15360MiB | 0% Default |
| | | N/A |
+-----------------------------------------+----------------------+----------------------+- Create a Workbench and clone the repo. Execute the llm_rest_requests.ipynb notebook to query the LLM model.