Update readme with new commands

zenml-io · Jan 21, 2024 · 8da2698 · 8da2698
1 parent 4b01ed5
commit 8da2698
Show file tree

Hide file tree

Showing 3 changed files with 7 additions and 6 deletions.
diff --git a/llm-finetuning/huggingface/README.md b/llm-finetuning/huggingface/README.md
@@ -3,12 +3,13 @@
 To register a custom huggingface flavor run the following
 
 ```bash
+zenml init
 zenml stack register zencoder_hf_stack -o default -a default
 zenml stack set zencoder_hf_stack
 export HUGGINGFACE_USERNAME=<here>
 export HUGGINGFACE_TOKEN=<here>
 zenml secret create huggingface_creds --username=$HUGGINGFACE_USERNAME --token=$HUGGINGFACE_TOKEN
-zenml model-deployer flavor register huggingface.hf_model_deployer_flavor.HFModelDeployerFlavor
+zenml model-deployer flavor register huggingface.hf_model_deployer_flavor.HuggingFaceModelDeployerFlavor
 ```
 
 Afterward, you should see the new flavor in the list of available flavors:
@@ -20,6 +21,6 @@ zenml model-deployer flavor list
 Register model deployer component into the current stack
 
 ```bash
-zenml model-deployer register hfendpoint --flavor=hfendpoint
+zenml model-deployer register hfendpoint --flavor=hfendpoint --token=$HUGGINGFACE_TOKEN
 zenml stack update zencoder_hf_stack -d hfendpoint
 ```
diff --git a/llm-finetuning/huggingface/hf_deployment.py b/llm-finetuning/huggingface/hf_deployment.py
@@ -27,7 +27,7 @@ class HuggingFaceDeploymentConfig(ServiceConfig):
     instance_type: str
     region: str
     vendor: str
-    hf_token: str
+    token: str
     account_id: Optional[str] = None
     min_replica: Optional[int] = 0
     max_replica: Optional[int] = 1
@@ -69,7 +69,7 @@ def hf_endpoint(self) -> InferenceEndpoint:
         """
         return get_inference_endpoint(
             name=self.config.endpoint_name,
-            token=self.config.hf_token,
+            token=self.config.token,
             namespace=self.config.namespace,
         )
 
@@ -93,7 +93,7 @@ def provision(self) -> None:
             custom_image=self.config.custom_image,
             type=self.config.endpoint_type,
             namespace=self.config.namespace,
-            token=self.config.hf_token,
+            token=self.config.token,
         ).wait(timeout=POLLING_TIMEOUT)
 
         if self.hf_endpoint.url is not None:

diff --git a/llm-finetuning/steps/deployment.py b/llm-finetuning/steps/deployment.py
@@ -106,7 +106,7 @@ def deploy_model_to_hf_hub(
         instance_type=instance_type,
         region=region,
         vendor=vendor,
-        hf_token=hf_token,
+        token=hf_token,
         account_id=account_id,
         min_replica=min_replica,
         max_replica=max_replica,