gollama

Gollama

Gollama is a tool for managing Ollama models.

It provides a TUI (Text User Interface) for listing, inspecting, deleting, copying, and pushing Ollama models as well as optionally linking them to LM Studio.

The application allows users to interactively select models, sort them by various criteria, and perform actions on them using hotkeys.

Table of Contents

• Gollama
  • Table of Contents
  • Features
  • Installation
  • Usage
    • Key Bindings
      • Command-line Options
  • Configuration
  • Installation and build from source
  • Logging
  • Contributing
  • License
  • Acknowledgements
  • Diagrams
    • Workflow
    • Components

Features

The project started off as a rewrite of my llamalink project, but I decided to expand it to include more features and make it more user-friendly.

It's in the early stages of development, so there are plenty of bugs and missing features, but I'm already finding it useful for managing my models, especially for cleaning up old models.

• Interactive TUI with sorting and filtering capabilities.
• List available models and display basic metadata such as size, quantization level, model family, and modified date.
• Sort models by name, size, modification date, quantization level, and family.
• Select and delete models.
• Inspect model for additional details.
• Link models to LM Studio.
• Copy models.
• Push models to a registry.
• Show running models.
• Plenty more comings soon if I continue to find the tool useful.

Installation

From go:

go install github.com/sammcj/gollama@latest

Or for bleeding edge:

go install github.com/sammcj/gollama@HEAD

From Github:

Download the most recent release from the releases page and extract the binary to a directory in your PATH.

e.g. zip -d gollama-v1.4.2.zip -d gollama && mv gollama /usr/local/bin

Usage

To run the gollama application, use the following command:

gollama

Key Bindings

• Space: Select
• Enter: Run model (Ollama run)
• i: Inspect model
• t: Top (show running models)
• D: Delete model
• c: Copy model
• u: Update model (edit Modelfile) Work in progress
• P: Push model
• n: Sort by name
• s: Sort by size
• m: Sort by modified
• k: Sort by quantization
• f: Sort by family
• l: Link model to LM Studio
• L: Link all models to LM Studio
• q: Quit

Command-line Options

• -l: List all available Ollama models and exit
• -ollama-dir: Custom Ollama models directory
• -lm-dir: Custom LM Studio models directory
• -no-cleanup: Don't cleanup broken symlinks
• -cleanup: Remove all symlinked models and empty directories and exit
• -v: Print the version and exit

Gollama can also be called with -l to list models without the TUI.

gollama -l

Configuration

Gollama uses a JSON configuration file located at ~/.config/gollama/config.json. The configuration file includes options for sorting, columns, API keys, log levels etc...

Example configuration:

{
  "default_sort": "modified",
  "columns": [
    "Name",
    "Size",
    "Quant",
    "Family",
    "Modified",
    "ID"
  ],
  "ollama_api_key": "",
  "ollama_api_url": "http://localhost:11434",
  "lm_studio_file_paths": "",
  "log_level": "info",
  "log_file_path": "/Users/username/.config/gollama/gollama.log",
  "sort_order": "Size",
  "strip_string": "my-private-registry.internal/",
  "editor": "",
  "docker_container": ""
}

• strip_string can be used to remove a prefix from model names as they are displayed in the TUI. This can be useful if you have a common prefix such as a private registry that you want to remove for display purposes.
• docker_container - experimental - if set, gollama will attempt to perform any run operations inside the specified container.
• editor - experimental - if set, gollama will use this editor to open the Modelfile for editing.

Installation and build from source

1. Clone the repository:

   git clone https://github.com/sammcj/gollama.git
   cd gollama
   

2. Build:

   go get
   make build
   

3. Run:

   ./gollama
   

Logging

Logs can be found in the gollama.log which is stored in $HOME/.config/gollama/gollama.log by default. The log level can be set in the configuration file.

Contributing

Contributions are welcome! Please fork the repository and create a pull request with your changes.

License

Copyright © 2024 Sam McLeod

This project is licensed under the MIT License. See the LICENSE file for details.

Acknowledgements

• Ollama
• Llama.cpp
• Charmbracelet

Thank you to folks such as Matt Williams for giving this a shot and providing feedback.

Diagrams

Workflow

sequenceDiagram
    participant User
    participant AppModel
    participant TextInputModel
    participant API

    User ->> AppModel: Select "Copy Model"
    AppModel ->> TextInputModel: Prompt for new name
    TextInputModel ->> User: Display text input
    User ->> TextInputModel: Enter new name
    TextInputModel ->> AppModel: Return new name
    AppModel ->> API: Copy model with new name
    API ->> AppModel: Confirm copy success
    AppModel ->> API: Push new model
    API ->> AppModel: Display progress bar
    API ->> AppModel: Confirm push success
    AppModel ->> User: Display success message

Components

graph TD
    A[Project Structure] --> B[main.go]
    A --> C[app_model.go]
    A --> D[config/config.go]
    A --> E[helpers.go]
    A --> F[item_delegate.go]
    A --> G[keymap.go]
    A --> H[logging/logging.go]
    A --> I[model.go]
    A --> J[operations.go]
    A --> K[styles.go]
    A --> L[Makefile]
    A --> N[top_view.go]
    A --> O[text_input.go]

    B --> |Initializes| C
    B --> |Loads| D
    B --> |Uses| G
    B --> |Uses| H
    B --> |Uses| J
    C --> |Manages| F
    C --> |Calls| E
    C --> |Uses| G
    C --> |Logs with| H
    C --> |Displays| I
    C --> |Applies| K
    J --> |Interacts with| D
    J --> |Logs with| H