Sold by: Upstage
Deployed on AWS
Free Trial
Solar Embedding Large: A multilingual model optimized for retrieval tasks in English, Korean, Japanese, and more.
Overview
Solar Embedding Large is a powerful multilingual embedding model offering robust performance across multiple languages, including English, Korean, Japanese, and more. It's specifically fine-tuned for retrieval tasks, significantly enhancing multilingual retrieval results. This model is divided into two specialized versions: 'solar-embedding-1-large-query', optimized for embedding user's question, and 'solar-embedding-1-large-passage', designed for embedding documents to be searched. Utilizing these purpose-specific models increases retrieval efficiency, which leads to improved performance of Retrieval Augmented Generation (RAG) systems.
Highlights
- ## Key Features - **Enhanced RAG System Efficiency:** The specialized models for query and passage embedding in Solar Embedding Large significantly improve the performance and precision of Retrieval Augmented Generation (RAG) systems. - **Robust Multilingual Capabilities:** This model demonstrates strong performance in multiple languages. It is exceptional choice for multilingual retrieval tasks, including English, Korean, Japanese, and more.
- ## Key Applications Solar Embedding Large excels in diverse tasks, particularly in Retrieval Augmented Generation (RAG), where precise information retrieval is crucial. Solar's query and passage models are optimized for these tasks. With robust language support including English, Korean, Japanese, and more, it is well-suited for a wide range of information retrieval applications requiring high accuracy across multiple languages.
- ## Key Tasks - Retrieval Augmented Generation - Semantic Search - Reranking - Text Embeddings
Details
Sold by
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
Free trial
Try this product free for 7 days according to the free trial terms set by the vendor.
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.g6.2xlarge Inference (Batch) Recommended | Model inference on the ml.g6.2xlarge instance type, batch mode | $0.80 |
ml.g6.2xlarge Inference (Real-Time) Recommended | Model inference on the ml.g6.2xlarge instance type, real-time mode | $0.80 |
ml.g6.4xlarge Inference (Batch) | Model inference on the ml.g6.4xlarge instance type, batch mode | $0.80 |
ml.g6.8xlarge Inference (Batch) | Model inference on the ml.g6.8xlarge instance type, batch mode | $0.80 |
ml.g6.12xlarge Inference (Batch) | Model inference on the ml.g6.12xlarge instance type, batch mode | $3.20 |
ml.g6.16xlarge Inference (Batch) | Model inference on the ml.g6.16xlarge instance type, batch mode | $0.80 |
ml.g6.24xlarge Inference (Batch) | Model inference on the ml.g6.24xlarge instance type, batch mode | $3.20 |
ml.g6.48xlarge Inference (Batch) | Model inference on the ml.g6.48xlarge instance type, batch mode | $6.40 |
ml.g6.4xlarge Inference (Real-Time) | Model inference on the ml.g6.4xlarge instance type, real-time mode | $0.80 |
ml.g6.8xlarge Inference (Real-Time) | Model inference on the ml.g6.8xlarge instance type, real-time mode | $0.80 |
Dimensions summary
You pay by the host-hour based on the AWS instance type you run. Pricing scales with instance size, from ml.g6.2xlarge up to ml.g6.48xlarge, so larger instances cost more per hour. You choose between two inference modes. Batch mode processes grouped workloads and is available on the g6 instance family. Real-time mode handles live requests and is offered on both the g6 and g6e instance families. The g6e family adds real-time options only. You are billed for the hours each instance runs, with no upfront commitment.
Top-of-mind questions for buyers
A host-hour is one hour that one instance of your chosen type runs the model. You are billed per hour the instance stays active. A fully stopped instance stops software charges, though underlying AWS storage fees may still apply while resources remain provisioned.
Both bill per host-hour on running instances. Batch mode processes grouped workloads and runs on the g6 family. Real-time mode handles live requests and runs on both the g6 and g6e families. You pick the mode and instance type; cost tracks the hours each instance runs.
Pricing scales with instance size, so moving from a smaller to a larger instance raises your per-hour rate. The change is not automatic; you deploy the instance type you choose. You are billed for the hours that specific instance runs, with no upfront commitment.
www.upstage.ai
Helpful?
Vendor refund policy
We do not support any refunds currently.
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 model
An Amazon SageMaker model package is a pre-trained machine learning model ready to use without additional training. Use the model package to create a 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
- Enhanced throughput by supporting larger instance types
- Added embedding dimension reduction support
- Introduced batch inference support
Additional details
Inputs
- Summary
We support the request payload compatible with OpenAI's Embeddings API endpoint.
- Input MIME type
- application/json, application/jsonlines
# Input is single string
input = {
"input": "How is the performance of Solar embeddings?",
"model": "solar-embedding-1-large-query"
}
# Input is a list of string
input = {
"input": [
"Solar embeddings are awesome.",
"Solar embedding large model demonstrates strong performance in multiple languages."
],
"model": "solar-embedding-1-large-passage"
}
{"model": "embedding-passage", "input": "Solar Embedding 1 Large is great!", "dimensions": 2048}
{"model": "embedding-passage", "input": "Upstage builds frontier AI for enterprises.", "dimensions": 2048}
{"model": "embedding-passage", "input": "Embedding is a key component when constructing RAG system", "dimensions": 2048}
Input data descriptions
The following table describes supported input data fields for real-time inference and batch transform.
Field name | Description | Constraints | Required |
|---|---|---|---|
input | A single text string, or an array of texts to embed. | - | Yes |
model | Name of the model utilized to carry out the embedding. Current available models are 'solar-embedding-1-large-query' and 'solar-embedding-1-large-passage'. | - | Yes |
dimensions | The number of dimensions for the output embedding vector. Allowed values are integers from 1 to 2048, or 4096. default: 4096 | - | No |
Resources
Vendor resources
Support
Vendor support
Contact us for model fine-tuning request.
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
Embed 5 is Cohere's most powerful embedding family yet, delivering frontier retrieval on complex enterprise data that matches or exceeds the competition and makes major gains on Embed 4.
Embed 5 is Cohere's most powerful embedding family yet, delivering frontier retrieval on complex enterprise data that matches or exceeds the competition and makes major gains on Embed 4.
Embed 4 transforms multiple modalities, such as texts and images, into numerical vectors.
Cohere provides a representative AI model, Embed Light, that translates texts and images into numerical vectors that models can understand.
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.