from PyAnnote.audio import Pipeline

def model_fn(model_dir):
# Load the model from the specified model directory
model = Pipeline.from_pretrained(
"PyAnnote/speaker-diarization-3.1",
use_auth_token="Replace-with-the-Hugging-face-auth-token")
return model

with open("model/code/requirements.txt", "w") as f:
    f.write("transformers==4.25.1n")
    f.write("boto3n")
    f.write("PyAnnote.audion")
    f.write("soundfilen")
    f.write("librosan")
    f.write("onnxruntimen")
    f.write("wgetn")
    f.write("pandas")

!tar zcvf model.tar.gz *

import sagemaker
import boto3

sess = sagemaker.Session()

sagemaker_session_bucket = None
if sagemaker_session_bucket is None and sess is not None:
    sagemaker_session_bucket = sess.default_bucket()

try:
    role = sagemaker.get_execution_role()
except ValueError:
    iam = boto3.client("iam")
    role = iam.get_role(RoleName="sagemaker_execution_role")["Role"]["Arn"]

sess = sagemaker.Session(default_bucket=sagemaker_session_bucket)

print(f"sagemaker role arn: {role}")
print(f"sagemaker bucket: {sess.default_bucket()}")
print(f"sagemaker session region: {sess.boto_region_name}")

s3_location = f"s3://{sagemaker_session_bucket}/whisper/model/model.tar.gz"
!aws s3 cp model.tar.gz $s3_location

from sagemaker.huggingface.model import HuggingFaceModel
from sagemaker.async_inference.async_inference_config import AsyncInferenceConfig
from sagemaker.s3 import s3_path_join
from sagemaker.utils import name_from_base

async_endpoint_name = name_from_base("custom-asyc")

# create Hugging Face Model Class
huggingface_model = HuggingFaceModel(
    model_data=s3_location,  # path to your model and script
    role=role,  # iam role with permissions to create an Endpoint
    transformers_version="4.17",  # transformers version used
    pytorch_version="1.10",  # pytorch version used
    py_version="py38",  # python version used
)

# create async endpoint configuration
async_config = AsyncInferenceConfig(
    output_path=s3_path_join(
        "s3://", sagemaker_session_bucket, "async_inference/output"
    ),  # Where our results will be stored
    # Add nofitication SNS if needed
    notification_config={
        # "SuccessTopic": "PUT YOUR SUCCESS SNS TOPIC ARN",
        # "ErrorTopic": "PUT YOUR ERROR SNS TOPIC ARN",
    },  #  Notification configuration
)

env = {"MODEL_SERVER_WORKERS": "2"}

# deploy the endpoint endpoint
async_predictor = huggingface_model.deploy(
    initial_instance_count=1,
    instance_type="ml.xx",
    async_inference_config=async_config,
    endpoint_name=async_endpoint_name,
    env=env,
)

# Replace with a path to audio object in S3
from sagemaker.async_inference import WaiterConfig
res = async_predictor.predict_async(data=data)
print(f"Response output path: {res.output_path}")
print("Start Polling to get response:")

config = WaiterConfig(
  max_attempts=10, #  number of attempts
  delay=10#  time in seconds to wait between attempts
  )
res.get_result(config)
#import waiterconfig

[0.9762308998, 8.9049235993, "SPEAKER_01"]

[9.533106961, 12.1646859083, "SPEAKER_01"]

[13.1324278438, 13.9303904924, "SPEAKER_00"]

[14.3548387097, 26.1884550085, "SPEAKER_00"]

[27.2410865874, 28.2258064516, "SPEAKER_01"]

[28.3446519525, 31.298811545, "SPEAKER_01"]

# Common class representing application autoscaling for SageMaker 
client = boto3.client('application-autoscaling') 

# This is the format in which application autoscaling references the endpoint
resource_id='endpoint/' + <endpoint_name> + '/variant/' + <'variant1'> 

# Define and register your endpoint variant
response = client.register_scalable_target(
    ServiceNamespace='sagemaker', 
    ResourceId=resource_id,
    ScalableDimension='sagemaker:variant:DesiredInstanceCount', # The number of EC2 instances for your Amazon SageMaker model endpoint variant.
    MinCapacity=0,
    MaxCapacity=5
)

Idan kana son share wurin ƙarshe, yi amfani da lambar mai zuwa:

async_predictor.delete_endpoint(async_endpoint_name)

Fa'idodin tura ƙarshen asynchronous

Wannan maganin yana ba da fa'idodi masu zuwa:

• Maganin zai iya sarrafa fayilolin mai jiwuwa da yawa ko manyan.
• Wannan misali yana amfani da misali guda ɗaya don nunawa. Idan kuna son amfani da wannan mafita don ɗaruruwan ko dubban bidiyoyi kuma kuyi amfani da ƙarshen ƙarshen asynchronous don aiwatarwa cikin yanayi da yawa, zaku iya amfani da tsarin sikelin atomatik, wanda aka tsara don babban adadin takardun tushe. Sikelin atomatik yana daidaita adadin lokuta da aka tanadar don ƙira don amsa canje-canje a cikin aikin ku.
• Maganin yana inganta kayan aiki kuma yana rage nauyin tsarin ta hanyar raba ayyuka masu tsawo daga ainihin lokaci.

Kammalawa

A cikin wannan sakon, mun samar da madaidaiciyar hanya don tura samfurin hugging Face's lasifikar magana akan SageMaker ta amfani da rubutun Python. Yin amfani da ƙarshen ƙarshen asynchronous yana ba da ingantacciyar hanya da ma'auni don sadar da tsinkayar tsinkaya azaman sabis, karɓar buƙatun lokaci guda ba tare da matsala ba.

Fara yau tare da asynchronous lasifikar don ayyukan sautin ku. Tuntuɓi a cikin sharhin idan kuna da wasu tambayoyi game da samun ƙarshen diarization ɗin ku asynchronous ƙarshe da aiki.

Game da Authors

Sanjay Tiwary Shin ƙwararrun masanin ƙwararrun masanan caca Ai / ml ne ke kashe lokacinsa aiki tare da abokan cinikin da zasu ayyana takamaiman kayan amfani, da kuma yin amfani da Ai / ml da sabis ɗin AI / ml, da kuma masu yawansu. Ya taimaka ƙaddamar da ƙaddamar da AI / ML mai amfani da sabis na Amazon SageMaker kuma ya aiwatar da hujjoji da yawa na ra'ayi ta amfani da sabis na Amazon AI. Har ila yau, ya haɓaka dandalin nazari na ci gaba a matsayin wani ɓangare na tafiyar canjin dijital.

Kiran Challapalli babban mai haɓaka kasuwancin fasaha ne tare da sashin jama'a na AWS. Yana da fiye da shekaru 8 na gwaninta a AI / ML da shekaru 23 na ci gaban software da ƙwarewar tallace-tallace. Kiran yana taimaka wa kamfanoni na jama'a a duk faɗin Indiya bincika da haɓaka hanyoyin samar da tushen girgije waɗanda ke amfani da AI, ML, da AI na haɓaka - gami da manyan ƙirar harshe — fasaha.