---
license: mit
tags:
- text-to-image
- lora
- diffusers
- template:sd-lora
- education
- childrens illustrations
- hand-drawn
- flux
base_model: black-forest-labs/FLUX.1-dev
instance_prompt: hand drawn illustration
widget:
- text: 'hand drawn illustration of an alien with white background'
  output:
    url: sida-hd (1).png
- text: 'hand drawn illustration of napoleon'
  output:
    url: sida-hd (10).png

---

# Sida Hand Drawn 0.1 (sida-hd-01)

<Gallery />

## Model description

Sida Hand Drawn 0.1 (sida-hd-01) is a LoRA (Low-Rank Adaptation) model fine-tuned on the FLUX.1-dev base model. It is specifically designed to generate hand-drawn style images for an upcoming educational game called "Learn Isle" targeted at children. This model was created by the inpocket.ai team for use in inpocket.games projects, featuring the distinctive hand-drawn style of artist Sida.

The model works best when you include 'hand drawn illustration of a' at the beginning of your prompt. It excels at creating illustrations of objects, people, and items in a child-friendly, educational context.

Training details:
- LoRA Rank: 32
- Training Steps: 3000
- Learning Rate: 1e-4
- Batch Size: 1
- Resolution: [512, 768, 1024]
- Scheduler: FlowMatch
- Optimizer: AdamW8bit
- Precision: bfloat16

The model was trained on a custom dataset of hand-drawn illustrations by artist Sida, specifically created for the "Learn Isle" educational game.

## Trigger words

You should use `hand drawn illustration of a` at the beginning of your prompt to trigger the image generation in the intended style.

## Download model

Weights for this model are available in Safetensors format.

[Download](/inpocketai/sida-hd-01/tree/main) them in the Files & versions tab.

## Use it with the [🧨 diffusers library](https://github.com/huggingface/diffusers)

```py
from diffusers import FluxPipeline, FluxLoraConfig
import torch

model_id = "black-forest-labs/FLUX.1-dev"
lora_id = "inpocketai/sida-hd-01"

pipe = FluxPipeline.from_pretrained(model_id, torch_dtype=torch.float16).to("cuda")
lora_config = FluxLoraConfig.from_pretrained(lora_id)
pipe.load_lora_weights(lora_id)

prompt = "hand drawn illustration of a friendly robot teaching mathematics"
image = pipe(prompt, num_inference_steps=20, guidance_scale=4.0).images[0]
image.save("sida_hd_math_robot.png")
```

For more details, including weighting, merging and fusing LoRAs, check the [documentation on loading LoRAs in diffusers](https://huggingface.co/docs/diffusers/main/en/using-diffusers/loading_adapters)

## Ethical Considerations

This model is designed for creating educational content for children. Users should ensure that generated content is appropriate, educational, and free from harmful or inappropriate elements. The model should not be used to generate content that could be misleading or detrimental to a child's learning experience.

## Additional Information

- Developer: inpocket.ai team for inpocket.games
- Version: 0.1
- Date Created: September 27, 2024
- License: MIT
- Artist: Sida
- HD: In the context of this model, HD stands for "Hand Drawn"

For more information about Learn Isle, inpocket.ai, and inpocket.games, please visit our [website](https://www.inpocket.ai) or contact our team.