Samba-1-Instruct-Router

The Samba-1-Instruct-Router is a Composition of Experts (CoE) router. It provides a single model experience to a subset of the Samba-1 experts that comprise the Enterprise Grade AI (EGAI) benchmark. The EGAI benchmark is a comprehensive collection of widely adapted benchmarks sourced from the open source community. Each benchmark is carefully selected to measure specific aspects of a model’s capability to perform tasks pertinent to enterprise applications and use cases. See Benchmarking Samba-1 External link for more details.

Tasks

The Samba-1-Instruct-Router can be used for a variety of instruct-based tasks accessed from a single endpoint. It excels at general tasks that instruct models perform, including: generating text across different styles and formats, retrieving detailed information, decision-making support, automating repetitive tasks, writing and content creation, analyzing data, assisting in educational settings, and aiding in programming.

The list below describes some of the tasks for which you can use the Samba-1-Instruct-Router. You can also view Benchmarking Samba-1 External link for more information on supported tasks and benchmark performance.

Example tasks

  • Tables and databases

    • SQL query generation

    • Table question answering

  • General programming tasks

  • Math

  • Tool selection and API usage

    • Tool selection and API usage

    • API function calling

  • Parameter identification

  • Writing and brainstorming

  • Guardrails via critiques of LLM responses for the following tasks

    • Summarization

    • Examination

    • Coding

  • Medicine, law, and finance domain tasks

Router attributes

The table below describes the experts and components available in the Samba-1-Instruct-Router.

Version Experts Component Mode

Samba-1-Instruct-Router-V1

  • e5-mistral-7b-instruct

  • tulu-2-dpo-70b

  • TableLlama

  • Xwin-Math-70B-V1.0

  • deepseek-coder-33b-instruct

  • SambaCoder-nsql-llama-2-70b

  • Falcon-40b-instruct

  • UniNER-7B-all

  • Explore-LM-7B-Rewriting

  • NexusRaven-V2-13B

  • autoj-13b

  • finance-chat

  • medicine-chat

  • law-chat

KNN-Classifier

Inference

Inference inputs

Inference inputs for the Samba-1-Instruct-Router are described below. See the Online generative inference section of the API document for request format details.

  • Number of inputs: 1

  • Input: Prompt or query

  • Input type: Text

  • Input format: String

Example input
["What are some common stocks?"]

Task-specific prompt templates

To get optimal responses from the Samba-1-Instruct-Router experts listed below, use a task-specific prompt template to format your input. Click the links below for each expert to navigate to the corresponding template.

Inference outputs

Inference outputs for the Samba-1-Instruct-Router are described below. See the Online generative inference section of the API document for request format details.

  • Number of outputs: 1

  • Output: Model response

  • Output type: Text

  • Output format: Tokens

Example output
[
     "\n",
     "\n",
     "Answer",
     ":",
     "The",
     "capital",
     "of",
     "Austria",
     "is",
     "Vienna",
     "(",
     "G",
     "erman",
     ":",
     "Wien",
     ")."
],

Inference hyperparameters and settings

The table below describes the inference hyperparameters and settings the Samba-1-Instruct-Router. See the Online generative inference section of the API document for basic parameter details.

Attribute Type Description Allowed values

process_prompt

Boolean

Allows for the user to specify the complete prompt construction. If this value is unset, the prompt construction elements such as the start and stop tag, bot and assistant tag, and default system prompt will be constructed automatically.

true, false
Default: false

return_token_count_only

Boolean

If set to true, the endpoint will not run completion and simply return the number.

true, false
Default: false

select_expert

String

Manually select which expert model to invoke.

  • e5-mistral-7b-instruct

  • tulu-2-dpo-70b

  • TableLlama

  • Xwin-Math-70B-V1.0

  • deepseek-coder-33b-instruct

  • SambaCoder-nsql-llama-2-70b

  • Falcon-40b-instruct

  • UniNER-7B-all

  • Explore-LM-7B-Rewriting

  • NexusRaven-V2-13B

  • autoj-13b

  • finance-chat

  • medicine-chat

  • law-chat

Samba-1-Instruct-Router in the Playground

The Playground provides an in-platform experience for generating predictions using the Samba-1-Instruct-Router.