There is error below when trying to deploy HF model to Amazon SageMaker.
Error:
RuntimeError: The expanded size of the tensor (22528) must match the existing size (1024) at non-singleton dimension 0. Target sizes: [22528, 8192]. Tensor sizes: [1024, 8192]
SageMaker Instance: ml.g4dn.2xlarge
Code to deploy HF model to Amazon SageMaker:
import os
import json
import sagemaker
import boto3
from sagemaker.huggingface import HuggingFaceModel, get_huggingface_llm_image_uri
try:
role = sagemaker.get_execution_role()
except ValueError:
iam = boto3.client('iam')
role = iam.get_role(RoleName='sagemaker_execution_role')['Role']['Arn']
hub = {
'HF_MODEL_ID': 'meta-llama/Llama-2-7b-chat-hf',
'SM_NUM_GPUS': json.dumps(1),
'HUGGING_FACE_HUB_TOKEN': <HF_TOKEN>
}
huggingface_model = HuggingFaceModel(
image_uri=get_huggingface_llm_image_uri("huggingface",version="0.8.2"),
env=hub,
role=role
)
predictor = huggingface_model.deploy(
initial_instance_count=1,
instance_type="ml.g4dn.2xlarge",
container_startup_health_check_timeout=1800
)