Mirrors
/
hf_text-generation-inference
mirror of https://github.com/huggingface/text-generation-inference.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Update aml deployment

This commit is contained in:
Olivier Dehaene 2022-10-17 10:39:59 +02:00
parent bcb53903b8
commit 00e6ce44b1
4 changed files with 9 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
aml/README.md
View File

@ -2,6 +2,7 @@
docker build . -t db4c2190dd824d1f950f5d1555fbadf0.azurecr.io/text-generation:0.1 docker build . -t db4c2190dd824d1f950f5d1555fbadf0.azurecr.io/text-generation:0.1
docker push 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-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 az ml online-deployment create -f deployment.yaml -g HuggingFace-BLOOM-ModelPage -w HuggingFace
``` ```

5
aml/deployment.yaml
View File

@ -1,9 +1,7 @@
$schema: https://azuremlschemas.azureedge.net/latest/managedOnlineDeployment.schema.json $schema: https://azuremlschemas.azureedge.net/latest/managedOnlineDeployment.schema.json
name: bloom-deployment name: bloom-deployment
endpoint_name: bloom-inference endpoint_name: bloom-inference
model: model: azureml:bloom:1
name: bloom
path: ./bloom
model_mount_path: /var/azureml-model model_mount_path: /var/azureml-model
environment_variables: environment_variables:
MODEL_BASE_PATH: /var/azureml-model/bloom MODEL_BASE_PATH: /var/azureml-model/bloom
@ -24,6 +22,7 @@ environment:
instance_type: Standard_ND96amsr_A100_v4 instance_type: Standard_ND96amsr_A100_v4
request_settings: request_settings:
request_timeout_ms: 90000 request_timeout_ms: 90000
max_concurrent_requests_per_instance: 256
liveness_probe: liveness_probe:
initial_delay: 300 initial_delay: 300
timeout: 20 timeout: 20

2
aml/endpoint.yaml
View File

@ -1,3 +1,3 @@
$schema: https://azuremlsdk2.blob.core.windows.net/latest/managedOnlineEndpoint.schema.json $schema: https://azuremlsdk2.blob.core.windows.net/latest/managedOnlineEndpoint.schema.json
name: bloom-inference name: bloom-inference
auth_mode: aml_token auth_mode: key

5
aml/model.yaml Normal file
View File

@ -0,0 +1,5 @@
$schema: https://azuremlschemas.azureedge.net/latest/model.schema.json
name: bloom
version: 1
path: ./bloom
type: custom_model