Select Active LLM Model

Overview

Flow ID: select-llm-model
Category: Model Management
Estimated Duration: 10-20 seconds (loading time)
User Role: All Users
Complexity: Simple

Purpose: Designate which uploaded model handles the chat generation. Switching models is a “heavy” operation requiring unloading the old one and loading the new one into RAM.

Trigger

What initiates this flow:

• User manually initiates

Specific trigger: Header Dropdown (Quick Switch) or Settings.

Step-by-Step Flow

Main Path (Happy Path)

Step 1: Open Selector

• User Action: Click Model name in the top header.
• System Response: Dropdown lists available models.

Step 2: Choose Model

• User Action: Click Llama-3-70B.
• System Response:
  • Dropdown closes.
  • Loading spinner appears (“Loading Model…”).
  • System frees RAM from old model.
  • Loads new model (can take 5-30s).

Step 3: Ready

• System Response: Spinner stops. “Model Ready”.

Error States & Recovery

Error 1: Load Failed

Cause: Model file corrupted or missing
User Experience: “Failed to load model.” Reverts to previous or “None”.
Recovery: Re-upload model.

Related Flows

• LLM Model Upload

Technical References

• src/actions/model-actions.js (loadModel)