Kabatubare/midjourney_sdxl_llama2_7b

Model Card for Your-Text-to-Image-Model-ID

Model Details

Model Description

• Developed by: Bronx2HelsinkiAI
• Model type: Text-to-Image Generation
• Language(s) (NLP): English
• License: Other

Training Details

The base model used for midjourney-sdxl-llama2-7b is guardrail/llama-2-7b-guanaco-instruct-sharded

Training Data

The model was trained on a dataset that combines 718 rows from the Kabatubare/midjurney dataset with a curated set of samples from the Stable-Diffusion-Prompts dataset, which originally contained 73,000 rows.

Data Sources:

• Kabatubare/midjurney
• A curated set of samples from Stable-Diffusion-Prompts
• Synthetic Data: The dataset also includes synthetic prompts generated using GPT-4.

Data Enrichment:

The model benefits from a dataset that was enriched with curated samples from the Stable Diffusion prompts. These curated samples were added to the existing Midjourney dataset, offering a more diverse range of creative text-to-image prompts.