Skip to content

Latest commit

 

History

History
70 lines (48 loc) · 2.08 KB

README.md

File metadata and controls

70 lines (48 loc) · 2.08 KB

llm-gateway

Cohere

Setup

This setup requires a Cohere account. If you do not have a Cohere account, you can create one here.

  1. Login to your Cohere account and select API keys to view your API keys.
  2. If you do not have an existing API key, select Create Trial key to create one.
  3. Within llm-gateway, set the COHERE_API_KEY environment variable to your API key.

Available Models

  • Command Light

OpenAI

Setup

This setup requires an OpenAI account. If you do not have an OpenAI account, you can create one here.

  1. Login to your OpenAI account and navigate to the API page.
  2. Select API keys to view your API keys.
  3. If you do not have an existing API key, select Create new secret key to create one.
  4. Within llm-gateway, set the OPENAI_API_KEY environment variable to your API key.

Available Models

  • GPT 3.5 Turbo
  • GPT 3.5 Turbo 16k
  • GPT 4

AWS Bedrock

Setup

This setup requires an Amazon Web Services (AWS) account. If you do not have an AWS account, you can create one here.

  1. Enable Amazon Bedrock on your AWS account using the AWS Console.
  2. Within Amazon Bedrock in the AWS Console, select Model access > Manage Model Access, and enable the LLM models you want to use.
    Pricing information for models offered on Amazon Bedrock can be found here.
  3. Within llm-gateway, set the AWS_REGION environment variable to the region in which Amazon Bedrock is enabled on your AWS account.
  4. Within llm-gateway, set the AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY environment variables to the access key ID and secret access key of the AWS account you enabled Amazon Bedrock on.

Available Models

AI21 Labs

  • Jurrasic-2 Ultra
  • Jurassic-2 Mid

Amazon

  • Titan Text Lite
  • Titan Text Express
  • Titan Text Embeddings

Anthropic

  • Claude 2.1
  • Claude 2.0
  • Claude 1.3
  • Claude Instant

Cohere

  • Command
  • Command Light
  • Embed - English
  • Embed - Multilingual

Meta

  • Llama-2-13b-chat
  • Llama-2-70b-chat