Amazon Sagemaker
Amazon SageMaker is a fully-managed platform that enables developers and data scientists to quickly and easily build, train, and deploy machine learning models at any scale. With Amazon SageMaker, all the barriers and complexity that typically slow down developers who want to use machine learning are removed. The service includes models that can be used together or independently to build, train, and deploy your machine learning models.
Label Inspector
By:
Latest Version:
v0.2
Find label errors in any classification dataset (text and tabular/CSV datasets supported)
Product Overview
Cleanlab builds AI solutions to assess data quality in messy real-world applications. Mislabeled data is common in classification, but we invented Confident Learning algorithms that automatically detect label errors in your dataset. Label Inspector runs these algorithms to estimate which examples are likely mislabeled in any classification dataset. Simply provide the data (labels + features) for a classification task, and state-of-the-art ML models will be trained to score the quality of your labels and flag which ones are likely incorrect. Label Inspector can identify mislabeled examples in any standard multi-class classification dataset (including features that are: text, numeric, or categorical — with missing values allowed). It returns a CSV file with a row for each example in your dataset, stating: whether it appears mislabeled, how likely the label is correct, plus an alternative suggested label. Documentation and examples: https://github.com/cleanlab/aws-marketplace/
Key Data
Version
By
Type
Algorithm
Highlights
Label Inspector works for any standard multi-class classification dataset (including features that are: text, numeric, or categorical — with missing values allowed). It trains state-of-the-art ML models to automatically detect which examples are mislabeled.
Documentation and example usage notebooks for the latest version are available here: https://github.com/cleanlab/aws-marketplace/
Label Inspector auto-trains a robust ML model to identify potential label errors in your dataset. After the training is completed, you can deploy this trained model to classify any new data that you get. If your new data has an accompanying labels column, Label Inspector will also identify any potential label errors in the new data.
Not quite sure what you’re looking for? AWS Marketplace can help you find the right solution for your use case. Contact us
Pricing Information
Use this tool to estimate the software and infrastructure costs based your configuration choices. Your usage and costs might be different from this estimate. They will be reflected on your monthly AWS billing reports.
Contact us to request contract pricing for this product.
Estimating your costs
Choose your region and launch option to see the pricing details. Then, modify the estimated price by choosing different instance types.
Version
Region
Software Pricing
Algorithm Training$10/hr
running on ml.m5.xlarge
Model Realtime Inference$0.001/inference
running on any instance
Model Batch Transform$5.00/hr
running on ml.m5.xlarge
Infrastructure PricingWith Amazon SageMaker, you pay only for what you use. Training and inference is billed by the second, with no minimum fees and no upfront commitments. Pricing within Amazon SageMaker is broken down by on-demand ML instances, ML storage, and fees for data processing in notebooks and inference instances.
Learn more about SageMaker pricing
With Amazon SageMaker, you pay only for what you use. Training and inference is billed by the second, with no minimum fees and no upfront commitments. Pricing within Amazon SageMaker is broken down by on-demand ML instances, ML storage, and fees for data processing in notebooks and inference instances.
Learn more about SageMaker pricing
SageMaker Algorithm Training$0.23/host/hr
running on ml.m5.xlarge
SageMaker Realtime Inference$0.23/host/hr
running on ml.m5.xlarge
SageMaker Batch Transform$0.23/host/hr
running on ml.m5.xlarge
Algorithm Training
For algorithm training in Amazon SageMaker, the software is priced based on hourly pricing that can vary by instance type. Additional infrastructure cost, taxes or fees may apply.InstanceType | Algorithm/hr | |
|---|---|---|
ml.p3.2xlarge | $20.00 | |
ml.p3.16xlarge | $20.00 | |
ml.m5.24xlarge | $20.00 | |
ml.m5.xlarge Vendor Recommended | $10.00 |
Usage Information
Training
This method works for any standard multi-class classification dataset stored in a table (CSV file), where: each row corresponds to an example, entries in the label column are discrete categories (i.e. classes), and subsequent columns contain: numeric, categorical, or text (arbitrary string) values. These other columns are features used as predictors of the label by ML, and can have missing values.
If your dataset has text, use a p*-instance so language models can be fine-tuned on GPU. If your dataset is big (over 100k rows), use a big instance: m5.24xlarge if no text, p3.16xlarge otherwise.
Channel specification
Fields marked with * are required
training
*Input modes: File
Content types: test/csv
Compression types: None
Hyperparameters
Fields marked with * are required
runtime
Parameter to specify training mode. Options include: high_accuracy, fast.
Type: Categorical
Tunable: No
Model input and output details
Input
Summary
Your data should be in a CSV file where the first column contains the class labels (remaining columns will be treated as predictive features). The first line of the CSV file should be a header containing column names for your data.
Ensure that the labels are categorical strings (not continuous numbers but discrete integers are ok), as only multi-class and binary classification datasets are supported. Other columns of data table contain: numeric, categorical, or text (arbitrary string) values.
Input MIME type
text/csvSample input data
Output
Summary
Label Inspector outputs a CSV with 4 columns:
is_label_issuecontains True/False values specifying whether each example is inferred to be mislabeledlabel_scorecontains quality scores between 0 and 1 estimating the likelihood that each example is correctly labeled (lower scores indicate noiser labels)given_labelcontains the original label for each example (same as the first column of your input data)predicted_labelcontains a label for each example predicted by our ML model
Output MIME type
text/csvSample output data
Sample notebook
End User License Agreement
By subscribing to this product you agree to terms and conditions outlined in the product End user License Agreement (EULA)
Support Information
Label Inspector
For questions/support, please email: support@cleanlab.ai. Free Trials and Subscription Plans available! Email us for more details.
Your email subject line must state that you are using Label Inspector in AWS Marketplace.
AWS Infrastructure
AWS Support is a one-on-one, fast-response support channel that is staffed 24x7x365 with experienced and technical support engineers. The service helps customers of all sizes and technical abilities to successfully utilize the products and features provided by Amazon Web Services.
Learn MoreRefund Policy
We do not currently support refunds, but you can cancel your subscription to the service at any time.
Customer Reviews
Financial Services
must have in your tool kit
Dec 14, 2024
What do you like best about the product?lot of functionalities to play around with . That helps
massively as we crunch data sets.very helpful community and support modelWhat do you dislike about
the product?difficult to use but very handy once you get hang of it. Docuemntation is imp... Read more
massively as we crunch data sets.very helpful community and support modelWhat do you dislike about
the product?difficult to use but very handy once you get hang of it. Docuemntation is imp... Read more
Farming
amazing starting point for dataset curation
Oct 24, 2024
What do you like best about the product?clean api, very easy integration tutorials, now is one of my
go-to when i am taking new challenge.What do you dislike about the product?might by costly to none
intensive ds teamsWhat problems is the product solving and how is that benefiting you?Agriculture
go-to when i am taking new challenge.What do you dislike about the product?might by costly to none
intensive ds teamsWhat problems is the product solving and how is that benefiting you?Agriculture
Matt H.
View allA Must Have Aid For Precise Language Data Annotation
Aug 18, 2024Verified purchase review from AWS Marketplace
What do you like best about the product?Cleanlab Studio’s big advantage lies in automating the
finding of mislabeled data, a game-changer for our AI projects. It boasts an easy-to-use interface and
strong algorithms that significantly reduce data cleaning time, thereby allowing our team... Read more
finding of mislabeled data, a game-changer for our AI projects. It boasts an easy-to-use interface and
strong algorithms that significantly reduce data cleaning time, thereby allowing our team... Read more
Write a review
Share your thoughts about this product.
Write a customer review