Skip to content

Commit

Permalink
document for huggingface(vllm) servingruntime for multi-node
Browse files Browse the repository at this point in the history 
Signed-off-by: jooho lee <[email protected]>
  • Loading branch information
@Jooho
Jooho committed Oct 10, 2024
1 parent 0aafa48 commit 4e7cca4
Show file tree
Hide file tree
Showing 3 changed files with 362 additions and 0 deletions.
288 changes: 288 additions & 0 deletions docs/modelserving/v1beta1/llm/huggingface/multi-node/README.md
View file
Original file line number Diff line number Diff line change
@@ -0,0 +1,288 @@
# Multi-node/Multi-GPU Inference with Hugging Face LLM Serving Runtime

This guide provides step-by-step instructions on setting up multi-node and multi-GPU inference using Hugging Face's LLM Serving Runtime. Before proceeding, please ensure you meet the following prerequisites and understand the limitations of this setup.

## Prerequisites

- Multi-node functionality is only supported in **RawDeployment** mode.
- **Auto-scaling is not available** for multi-node setups. The auto-scaler will be automatically set to external.
- A **Persistent Volume Claim (PVC)** is required for multi-node configurations, and it must support the **ReadWriteMany (RWM)** access mode.


!!! note

You must have **exactly one head pod** in your setup. The replica count for this head pod can be adjusted using the `min_replicas` or `max_replicas` settings in the `InferenceService (ISVC)`. However, creating additional head pods will cause them to be excluded from the Ray cluster, resulting in improper functioning. Ensure this limitation is clearly documented.

## WorkerSpec and ServingRuntime

To enable multi-node/multi-GPU inference, a new API field, `workerSpec`, must be configured. The `huggingface-server-multinode` `ServingRuntime` includes this field and is built on **vLLM**, which supports multi-node/multi-GPU setups. Note that this setup is **not compatible with Triton**.

## Key Configurations

When using the `huggingface-server-multinode` `ServingRuntime`, there are two critical configurations you need to understand:

1. **`tensor-parallel-size`**:
This setting controls how many GPUs are used per node. You can configure it via the environment variable `"TENSOR_PARALLEL_SIZE"`. Once set, the GPU count (`nvidia.com/gpu`) in both the head and worker node deployment resources will be updated automatically.


2. **`pipeline-parallel-size`**:
This setting determines how many nodes are involved in the deployment. There are two ways to configure it:
- Set the `WorkerSpec.Size` field.
- Use the environment variable `"PIPELINE_PARALLEL_SIZE"`. Unlike `WorkerSpec.Size`, this variable represents the total number of nodes, including both the head and worker nodes.

Therefore, the `"PIPELINE_PARALLEL_SIZE"` should always be **1 greater** than `WorkerSpec.Size`. If both the `WorkerSpec.Size` and `"PIPELINE_PARALLEL_SIZE"` are specified, the **environment variable takes precedence**. The default value for `"PIPELINE_PARALLEL_SIZE"` is 2.


### Example Configuration

Here’s an example of an `InferenceService` configuration for a Hugging Face model:

```yaml
apiVersion: serving.kserve.io/v1beta1
kind: InferenceService
metadata:
name: huggingface-llama3
spec:
predictor:
model:
modelFormat:
name: huggingface
storageUri: pvc://llama-3-8b-pvc/hf/8b_instruction_tuned
env:
- name: TENSOR_PARALLEL_SIZE
value: "1"
- name: PIPELINE_PARALLEL_SIZE
value: "2"
resources:
limits:
cpu: "6"
memory: 24Gi
requests:
cpu: "6"
memory: 24Gi
workerSpec:
size: 1 # This will be ignored because PIPELINE_PARALLEL_SIZE is set in environments
```
## Serve the Hugging Face LLM Model Using 2 Nodes
Follow these steps to serve the Hugging Face LLM model using a multi-node setup.
### 1. Create a Persistent Volume Claim (PVC)
First, create a PVC for model storage. Be sure to update `%fileStorageClassName%` with your actual storage class.

```yaml
kubectl apply -f - <<EOF
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
name: llama-3-8b-pvc
spec:
accessModes:
- ReadWriteMany
volumeMode: Filesystem
resources:
requests:
storage: 50Gi
storageClassName: %fileStorageClassName%
EOF
```

### 2. Download the Model to the PVC

To download the model, export your Hugging Face token (`HF_TEST_TOKEN`) as an environment variable. Replace `%token%` with your actual token.


```yaml
export HF_TEST_TOKEN=%token%
export MODEL=meta-llama/Meta-Llama-3-8B-Instruct
curl -o download-model-to-pvc.yaml https://kserve.github.io/website/latest/modelserving/v1beta1/llm/huggingface/multi-node/download-model-to-pvc.yaml
envsubst < download-model-to-pvc.yaml | kubectl create -f -
```


### 3. Create a ServingRuntime

Apply the ServingRuntime configuration. Replace %TBD% with the path to your ServingRuntime YAML.

```bash
kubectl apply -f https://github.com/kserve/kserve/blob/master/config/runtimes/kserve-huggingfaceserver-multinode.yaml
```

### 4. Deploy the model

Finally, deploy the model using the following InferenceService configuration:

```yaml
kubectl apply -f - <<EOF
apiVersion: serving.kserve.io/v1beta1
kind: InferenceService
metadata:
name: huggingface-llama3
spec:
predictor:
model:
runtime: kserve-huggingfaceserver-multinode
modelFormat:
name: huggingface
storageUri: pvc://llama-3-8b-pvc/hf/8b_instruction_tuned
EOF
```

## Check `InferenceService` status.

To verify the status of your `InferenceService`, run the following command:

```bash
kubectl get inferenceservices huggingface-llama3
```

!!! success "Expected Output"
```{ .bash .no-copy }
NAME URL READY PREV LATEST PREVROLLEDOUTREVISION LATESTREADYREVISION AGE
huggingface-llama3 http://huggingface-llama3.default.example.com 12m
```



## Check `GPU resource` status.

To check the GPU resource status, follow these steps:

1. Retrieve the pod names for the head and worker nodes:
```bash
# Get pod name
podName=$(kubectl get pod -l app=isvc.huggingface-llama3-predictor --no-headers|cut -d' ' -f1)
workerPodName=$(kubectl get pod -l app=isvc.huggingface-llama3-predictor-worker --no-headers|cut -d' ' -f1)
```

2. Check the GPU memory size for both the head and worker pods:
```bash
# Check GPU memory size
kubectl exec $podName -- nvidia-smi
kubectl exec $workerPodName -- nvidia-smi
```

!!! success "Expected Output"
```{ .bash .no-copy }
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 550.90.07 Driver Version: 550.90.07 CUDA Version: 12.4 |
|-----------------------------------------+------------------------+----------------------+
| 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 NVIDIA A10G On | 00000000:00:1E.0 Off | 0 |
| 0% 33C P0 71W / 300W | 19031MiB / 23028MiB | 0% Default |
| | | N/A |
+-----------------------------------------+------------------------+----------------------+
...
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 550.90.07 Driver Version: 550.90.07 CUDA Version: 12.4 |
|-----------------------------------------+------------------------+----------------------+
| 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 NVIDIA A10G On | 00000000:00:1E.0 Off | 0 |
| 0% 30C P0 69W / 300W | 18959MiB / 23028MiB | 0% Default |
| | | N/A |
+-----------------------------------------+------------------------+----------------------+
```

## Perform Model Inference

You can perform model inference by forwarding the port for testing purposes. Use the following command:

```bash
kubectl port-forward pod/$podName 8080:8080
```

The KServe Hugging Face vLLM runtime supports the following OpenAI endpoints for inference:

- `/v1/completions`
- `/v1/chat/completions`

#### Sample OpenAI Completions request:

To make a request to the OpenAI completions endpoint, use the following `curl` command:

```bash
curl http://localhost:8080/openai/v1/completions \
-H "Content-Type: application/json" \
-d '{
"model": "huggingface-llama3",
"prompt": "At what temperature does Nitrogen boil?",
"max_tokens": 100,
"temperature": 0
}'
```
!!! success "Expected Output"
```{ .json .no-copy }
{
"id": "cmpl-3bf2b04bac4a43548bc657c999e4fe5f",
"choices": [
{
"finish_reason": "length",
"index": 0,
"logprobs": null,
"text": " Nitrogen is a colorless, odorless, tasteless, and non-toxic gas. It is a member of the group 15 elements in the periodic table. Nitrogen is a very common element in the universe and is found in many compounds, including ammonia, nitric acid, and nitrate salts.\nThe boiling point of nitrogen is -195.8°C (-320.4°F) at standard atmospheric pressure. This means that at this temperature, nitrogen will change from a liquid to a"
}
],
"created": 1728348255,
"model": "huggingface-llama3",
"system_fingerprint": null,
"object": "text_completion",
"usage": {
"completion_tokens": 100,
"prompt_tokens": 9,
"total_tokens": 109
}
}
```

#### Sample OpenAI Chat request:

To make a request to the OpenAI chat completions endpoint, use the following `curl` command:

```bash
curl http://localhost:8080/openai/v1/chat/completions \
-H "Content-Type: application/json" \
-d '{
"model": "huggingface-llama3",
"messages":[{"role":"system","content":"At what temperature does Nitrogen boil?"}],
"max_tokens": 100,
"stream": false
}'
```

!!! success "Expected Output"
```{ .json .no-copy }
{
"id": "cmpl-1201662139294b81b02aca115d7981b7",
"choices": [
{
"finish_reason": "stop",
"index": 0,
"message": {
"content": "The boiling point of nitrogen is -195.8°C (-320.44°F) at a pressure of 1 atm.",
"tool_calls": null,
"role": "assistant",
"function_call": null
},
"logprobs": null
}
],
"created": 1728348754,
"model": "huggingface-llama3",
"system_fingerprint": null,
"object": "chat.completion",
"usage": {
"completion_tokens": 26,
"prompt_tokens": 19,
"total_tokens": 45
}
}
```
53 changes: 53 additions & 0 deletions docs/modelserving/v1beta1/llm/huggingface/multi-node/download-model-to-pvc.yaml
View file
Original file line number Diff line number Diff line change
@@ -0,0 +1,53 @@
apiVersion: v1
kind: Pod
metadata:
name: setup-llama3-8b-binary
spec:
volumes:
- name: model-volume
persistentVolumeClaim:
claimName: llama-3-8b-pvc
restartPolicy: Never
initContainers:
- name: fix-volume-permissions
image: quay.io/quay/busybox:latest
imagePullPolicy: IfNotPresent
securityContext:
allowPrivilegeEscalation: true
resources:
requests:
memory: "64Mi"
cpu: "250m"
limits:
memory: "128Mi"
cpu: "500m"
command: ["sh"]
args: ["-c", "chown -R 1001:1001 /mnt/models"]
volumeMounts:
- mountPath: "/mnt/models/"
name: model-volume
containers:
- name: download-model
image: registry.access.redhat.com/ubi9/python-311:latest
imagePullPolicy: IfNotPresent
securityContext:
allowPrivilegeEscalation: true
resources:
requests:
memory: "1Gi"
cpu: "1"
limits:
memory: "1Gi"
cpu: "1"
command: ["sh"]
args:
[
"-c",
"pip install --upgrade pip && pip install --upgrade huggingface_hub && python3 -c 'from huggingface_hub import snapshot_download\nsnapshot_download(\n repo_id=\"${MODEL}\",\nlocal_dir=\"/mnt/models/hf/8b_instruction_tuned\",local_dir_use_symlinks=False,use_auth_token=\"${HF_TEST_TOKEN}\")'",
]
volumeMounts:
- mountPath: "/mnt/models/"
name: model-volume
env:
- name: TRANSFORMERS_CACHE
value: /tmp
21 changes: 21 additions & 0 deletions docs/modelserving/v1beta1/llm/huggingface/multi-node/vllm_runtime.yaml
View file
Original file line number Diff line number Diff line change
@@ -0,0 +1,21 @@
apiVersion: serving.kserve.io/v1beta1
kind: InferenceService
metadata:
name: huggingface-llama3
spec:
predictor:
model:
modelFormat:
name: huggingface
args:
- --model_name=llama3
- --model_id=meta-llama/meta-llama-3-8b-instruct
resources:
limits:
cpu: "6"
memory: 24Gi
nvidia.com/gpu: "1"
requests:
cpu: "6"
memory: 24Gi
nvidia.com/gpu: "1"

0 comments on commit 4e7cca4

Please sign in to comment.