Sold by: Mindbeam AI
Deployed on AWS
Litespark allows users to pretrain LLMs on AWS SageMaker from scratch with custom or Hugging Face datasets
Overview
Mindbeam's LLM Pre-training solution, Litespark, enables customers to pre-train their LLM models on AWS SageMaker from scratch using either a custom dataset or Hugging Face open-source dataset.
Highlights
- Pre-train your custom LLM in less than two days using 16 NVIDIA H100 GPUs (p5.48xlarge instances)
- Supports both open-source datasets on Hugging Face and custom datasets on Amazon EFS
- Efficient LLM Pre-training
Details
Sold by
Categories
Delivery method
Latest version
Deployed on AWS
New
Introducing multi-product solutions
You can now purchase comprehensive solutions tailored to use cases and industries.
Features and programs
Financing for AWS Marketplace purchases
AWS Marketplace now accepts line of credit payments through the PNC Vendor Finance program. This program is available to select AWS customers in the US, excluding NV, NC, ND, TN, & VT.
Pricing
Pricing is based on actual usage, with charges varying according to how much you consume. Subscriptions have no end date and may be canceled any time.
Additional AWS infrastructure costs may apply. Use the AWS Pricing Calculator to estimate your infrastructure costs.
Dimension | Description | Cost/host/hour |
|---|---|---|
ml.p3.2xlarge Inference (Batch) Recommended | Model inference on the ml.p3.2xlarge instance type, batch mode | $48.00 |
ml.p3.2xlarge Inference (Real-Time) Recommended | Model inference on the ml.p3.2xlarge instance type, real-time mode | $48.00 |
ml.p5.48xlarge Training Recommended | Algorithm training on the ml.p5.48xlarge instance type | $48.00 |
ml.p4de.24xlarge Training | Algorithm training on the ml.p4de.24xlarge instance type | $48.00 |
ml.p4d.24xlarge Training | Algorithm training on the ml.p4d.24xlarge instance type | $48.00 |
ml.p5e.48xlarge Training | Algorithm training on the ml.p5e.48xlarge instance type | $48.00 |
ml.p5en.48xlarge Training | Algorithm training on the ml.p5en.48xlarge instance type | $48.00 |
Dimensions summary
You pay by the host hour, billed on the AWS instance type you run. Pricing splits into two activities: inference and training. For inference, you choose the ml.p3.2xlarge instance in either batch mode or real-time mode. For training, you pick among five GPU instance types: ml.p4d.24xlarge, ml.p4de.24xlarge, ml.p5.48xlarge, ml.p5e.48xlarge, and ml.p5en.48xlarge. Each training instance carries its own hourly rate, so cost scales with the instance you select and the hours you use. There are no upfront commitments.
Top-of-mind questions for buyers
One host hour is one running hour of the specified AWS instance type. You are billed for each hour that instance runs the software. The ml.p3.2xlarge serves inference, while the ml.p4d, ml.p4de, ml.p5, ml.p5e, and ml.p5en instances handle training. Each instance type meters its own hours separately.
The software charge meters running host hours only. Instances you stop do not accrue host-hour software charges. Underlying AWS compute or storage fees may still apply while resources are provisioned, but the Litespark software bills for active running time on the chosen instance type.
Both run model inference on the same ml.p3.2xlarge instance and bill by host hour. Batch mode processes grouped requests, suited to scheduled or bulk workloads. Real-time mode serves live requests as they arrive. You pick the mode that matches your workload, and each carries its own hourly rate.
mindbeam.ai
Helpful?
Vendor refund policy
No refunds are available, please get in touch with support at support@mindbeam.ai for assistance.
Legal
Vendor terms and conditions
Upon subscribing to this product, you must acknowledge and agree to the terms and conditions outlined in the vendor's End User License Agreement (EULA) .
Content disclaimer
Vendors are responsible for their product descriptions and other product content. AWS does not warrant that vendors' product descriptions or other product content are accurate, complete, reliable, current, or error-free.
Delivery details
Amazon SageMaker algorithm
An Amazon SageMaker algorithm is a machine learning model that requires your training data to make predictions. Use the included training algorithm to generate your unique model artifact. Then deploy the model on Amazon SageMaker for real-time inference or batch processing. Amazon SageMaker is a fully managed platform for building, training, and deploying machine learning models at scale.
Deploy the model on Amazon SageMaker AI using the following options:
Version release notes
Updated performance, stable version.
Additional details
Inputs
- Summary
See Github for more information.
https://github.com/Mindbeam-AI/litespark-aws-marketplace/blob/main/realtime_sample_input.json
https://github.com/Mindbeam-AI/litespark-aws-marketplace/blob/main/batch_job_sample_input.json
Resources
Vendor resources
Support
Vendor support
For support, kindly send an email to support@mindbeam.ai
AWS infrastructure support
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.
Similar products
REQUIRES PRIVATE OFFER
To purchase LiteLLM Enterprise Self-Hosted, please reach out to sales@berri.ai for a Private Offer.
LiteLLM is an OpenAI compatible Proxy Server (LLM Gateway) to call 2,000+ LLM APIs using the OpenAI format Bedrock, Huggingface, VertexAI, TogetherAI, Azure OpenAI, OpenAI, etc. Get started with Opensource LiteLLM here: https://github.com/BerriAI/litellm (40,000+ Github Stars)
By Arcee
Llama-Spark is a conversational model developed by Arcee.ai. It excels at engaging in natural and informative conversations.
For customers interested in optimizing their workloads on AWS, we clarify how Mindbeam's solutions accelerate the pre-training, fine-tuning, and inference of large language models (LLMs) by leveraging proprietary algorithms to optimize GPU usage.
State-of-the-Art Natural Language Processing libraries and Python notebooks. Includes licensed software & models for text mining, DL and Visual model training, tuning, and testing.
Customer reviews
No customer reviews yet
Be the first to review this product . We've partnered with PeerSpot to gather customer feedback. You can share your experience by writing or recording a review, or scheduling a call with a PeerSpot analyst.