metadata
license: mit
base_model: hp-l33/aim-xlarge
tags:
- mamba
- lora
- image-generation
- simpsons
pipeline_tag: unconditional-image-generation
NicolasNoya/SimpsonsMamba
LoRA fine-tuned Mamba/AiM-based cartoon image generation model.
Model Details
- Base model: hp-l33/aim-xlarge (From https://arxiv.org/pdf/2408.12245)
- Fine-tuning method: LoRA
- Dataset: Simpsons cartoons
- Checkpoint step: 50500
Intended Use
This model is intended for research and educational image generation experiments.
Example Usage
Use the local generation script in this project with the corresponding checkpoint weights.
Training Notes
- Trained with periodic checkpointing and evaluation.
- This card was auto-generated and should be updated with detailed metrics and limitations.
Limitations
Outputs may reflect biases or artifacts in the training data.