Update aml deployment
This commit is contained in:
parent
bcb53903b8
commit
00e6ce44b1
|
@ -2,6 +2,7 @@
|
|||
docker build . -t db4c2190dd824d1f950f5d1555fbadf0.azurecr.io/text-generation:0.1
|
||||
docker push db4c2190dd824d1f950f5d1555fbadf0.azurecr.io/text-generation:0.1
|
||||
|
||||
az ml model create -f model.yaml -g HuggingFace-BLOOM-ModelPage -w HuggingFace
|
||||
az ml online-endpoint create -f endpoint.yaml -g HuggingFace-BLOOM-ModelPage -w HuggingFace
|
||||
az ml online-deployment create -f deployment.yaml -g HuggingFace-BLOOM-ModelPage -w HuggingFace
|
||||
```
|
|
@ -1,9 +1,7 @@
|
|||
$schema: https://azuremlschemas.azureedge.net/latest/managedOnlineDeployment.schema.json
|
||||
name: bloom-deployment
|
||||
endpoint_name: bloom-inference
|
||||
model:
|
||||
name: bloom
|
||||
path: ./bloom
|
||||
model: azureml:bloom:1
|
||||
model_mount_path: /var/azureml-model
|
||||
environment_variables:
|
||||
MODEL_BASE_PATH: /var/azureml-model/bloom
|
||||
|
@ -24,6 +22,7 @@ environment:
|
|||
instance_type: Standard_ND96amsr_A100_v4
|
||||
request_settings:
|
||||
request_timeout_ms: 90000
|
||||
max_concurrent_requests_per_instance: 256
|
||||
liveness_probe:
|
||||
initial_delay: 300
|
||||
timeout: 20
|
||||
|
|
|
@ -1,3 +1,3 @@
|
|||
$schema: https://azuremlsdk2.blob.core.windows.net/latest/managedOnlineEndpoint.schema.json
|
||||
name: bloom-inference
|
||||
auth_mode: aml_token
|
||||
auth_mode: key
|
||||
|
|
|
@ -0,0 +1,5 @@
|
|||
$schema: https://azuremlschemas.azureedge.net/latest/model.schema.json
|
||||
name: bloom
|
||||
version: 1
|
||||
path: ./bloom
|
||||
type: custom_model
|
Loading…
Reference in New Issue