Chat

Large Language Models (LLMs) like GPT-n from OpenAI, or Llama from Meta have been trained to generate text. We can use versions of these models that have been fine-tuned to act as chatbots, allowing you to interact with said models in a conversational style.

See also: Completion

Gradio pipeline

Text Generation

Step-by-step instructions to interact through your web browser with a HuggingFace model, running on Lyra, JupyterHub, rVDI, or your own machine.

Preset configurations

• Mixtral-8x7B-Instruct on Lyra with an A100 GPU
• Llama-2-7b-chat on Lyra with an A100 GPU
• Llama-2-7b-chat locally
• Bloom on Lyra with an A100 GPU

Custom configuration

Choose the model, service, and settings.

text-generation-webui

text-generation-webui is a popular web interface to interact with Large Language Models.

Ollama

Ollama is a CLI tool for interacting with Large Language Models.