Home › Training Tools › lucataco/ai-toolkit

lucataco/ai-toolkit - AI Training Tools Tool

Overview

lucataco/ai-toolkit is a Cog-packaged implementation of ostris/ai-toolkit published on Replicate and intended for training LoRA adapters for the FLUX.1-dev model family. It is designed to run training workflows from a reproducible Cog container, accepting a custom image dataset to produce LoRA checkpoints that can be attached to FLUX.1-dev-style models. The implementation focuses on dataset-driven fine-tuning (image-to-LoRA) rather than end-to-end model re-training. The repository is explicitly marked as deprecated in favor of ostris/flux-dev-lora-trainer on its Replicate listing, so users should treat lucataco/ai-toolkit as a legacy packaging of the original toolkit. According to the Replicate model page, lucataco/ai-toolkit implements the ostris/ai-toolkit training flow via Cog but will not be the recommended path for new training runs; the upstream ostris/flux-dev-lora-trainer is cited as the preferred replacement (see model page). Because this package is primarily a Cog wrapper, it is most useful when you need a containerized, reproducible training command for LoRA adapters and have an existing FLUX.1-dev-compatible dataset and downstream workflow.

Key Features

  • Cog-packaged implementation of ostris/ai-toolkit for reproducible, containerized training
  • Specifically targeted at creating LoRA adapters for the FLUX.1-dev model family
  • Accepts a custom image dataset to produce LoRA checkpoints for downstream use
  • Provides a reproducible training entrypoint suitable for Replicate-hosted execution
  • Marked deprecated; upstream ostris/flux-dev-lora-trainer is recommended for new runs

Example Usage

Example (python):

import replicate

# Example pattern for invoking a Replicate-hosted model. Replace inputs with names/values
# shown on the model page (https://replicate.com/lucataco/ai-toolkit).

client = replicate.Client()
model = client.models.get("lucataco/ai-toolkit")

# The actual input names and parameters depend on the model's API shown on Replicate.
# Use the model page to see required fields such as dataset path, output directory,
# training steps, and hyperparameters.

prediction = model.predict(
    dataset_path="/path/to/your/images.zip",  # placeholder: use the model's expected input
    train_steps=1000,
    learning_rate=1e-4,
    output_dir="/output/lora-checkpoint"
)

print(prediction)

# Note: lucataco/ai-toolkit is deprecated in favor of ostris/flux-dev-lora-trainer; check
# the replicate model page for the authoritative input schema and recommended trainer.
Last Refreshed: 2026-01-09

Key Information

  • Category: Training Tools
  • Type: AI Training Tools Tool
Visit Official Website

Related Tools

Hugging Face Accelerate

A simple way to launch, train, and use PyTorch models on almost any device with support for distributed configuration, automatic mixed precision (including fp8), and easy-to-configure FSDP/DeepSpeed.
Unsloth AI

Unsloth AI is an enterprise platform that accelerates fine-tuning of large language models and vision models by leveraging innovative quantization techniques. It enables faster performance (up to 2.2x faster) and uses significantly less VRAM, making model deployment and training more efficient. The organization also offers open-source tools and models, and is integrated with Hugging Face, with additional details available on its website.
AutoTrain

Hugging Face AutoTrain is an automated machine learning (AutoML) tool that allows users to train, evaluate, and deploy state-of-the-art ML models without writing code. It supports a range of tasks including text classification, image classification, token classification, summarization, question answering, translation, tabular data tasks, and LLM finetuning, with seamless integration into the Hugging Face ecosystem.
AI Toolkit (ostris)

All-in-one training suite for diffusion models (e.g., FLUX.1) with a UI and Modal/Docker workflows; supports LoRA/LoKr and conv training.
ostris/ai-toolkit

A GitHub repository offering a collection of AI scripts primarily for Stable Diffusion and related AI model training. It includes a web UI for managing and monitoring jobs as well as tools for training models like FLUX.1-dev.
DeepScaleR

DeepScaleR is an open-source project that democratizes reinforcement learning (RL) for large language models (LLMs). The repository provides training scripts, model checkpoints, detailed hyperparameter configurations, datasets, and evaluation logs to reproduce and scale RL techniques on LLMs, aimed at reproducibility and research in advanced AI training.