LogoLogo
  • Introduction
    • 🤔Rationale
  • Using the platform
    • 📝Creating your account
    • 📀Selecting an LLM
    • 💬Interacting with a character
    • 🐣Creating your first character
    • 🗃️Uploading data to your character
    • 🫂Sharing
    • 🔧Setting Up Bittensor
  • Resources
    • 🔗Directory
Powered by GitBook
On this page
  • Introduction
  • Default Selection
  • Step 1: Access Model Select
  • Step 2: Select Model
  • Centralized, Proprietary, and Closed Source LLMs
  • Decentralized and Open Source LLMs
  • Usage Policy
  1. Using the platform

Selecting an LLM

Last updated 1 year ago

Introduction

A core feature of the platform, users are able to seamlessly switch between different LLMs allowing for maximum flexibility in both character design and in the freedom to express yourself. Chat history and files are preserved across all models. As per the beta, the available LLMs are GPT-4 Turbo, Llama 2, Bittensor, and Gemini. Certainly by Phase 4 and possibly following Phase 2, users will be able to upload their own LLMs securely, either for their own private use or to share and monetize publicly.

Note: It is against our terms of service to use our platform to violate the terms of another service. We employ service-specific moderation to help prevent misuse. Deliberate attempts to circumvent moderation is strictly prohibited.

Default Selection

By default, all models on VectorChat run on GPT-4 Turbo. This offers a high-quality and great introductory conversational experience, although users must abide by the terms of use provided by OpenAI.

Step 1: Access Model Select

To change the active LLM, start by navigating to your Dashboard. Once there, click on the robot icon in the top right corner of the screen, next to your profile picture.

Step 2: Select Model

This will bring up a dropdown of the available LLMs on the VectorChat platform. Simply select the model you wish to use and it will become active. Selecting Bittensor requires a Corcel API key. For information on how to obtain this key and set up Bittensor, refer to following section of the documentation.

Centralized, Proprietary, and Closed Source LLMs

For the highest quality experience, using GPT-4 Turbo or Gemini is likely the best way to go—provided your use is within the terms of either OpenAI or Google respectively. Additionally, as per the beta, multi-modality is only available for GPT-4 Turbo and Gemini, although it will be released for Bittensor at some point down the road.

Decentralized and Open Source LLMs

For maximal flexibility in expression, decentralized and open source LLMs are an excellent option. Note that some open source LLMs may still have terms of use. Regardless of the terms or lack thereof of a model, all users must still abide by our terms.

Usage Policy

VectorChat strictly prohibits the use of our platform for:

  • CSAM

  • Explicit and actionable instruction on self-harm

  • Explicit and actionable instruction on criminal activity

  • Circumventing the terms of another service

📀
🔧Setting Up Bittensor
image of model select dropdown