abedalswaity7 / flux-prompt

3B

6 Pulls Updated 3 hours ago

19 hours ago

7beb96bb4bc0 · 3.4GB

model
llama
·
3.21B
·
Q8_0
template
Below are some instructions that describe some tasks. Write responses that appropriately complete each request.{{ if .Prompt }} ### Instruction: {{ .Prompt }}{{ end }} ### Response: {{ .Response }}<|end_of_text|>
params
{"stop":["<|eot_id|>","<|start_header_id|>","<|eom_id|>","<|end_header_id|>","<|python_tag|>","<|finetune_right_pad_id|>","<|end_of_text|>","<|reserved_special_token_"]}

Readme

I’ve been working on a powerful tool that connects language models with image generation using cutting-edge tech. Here’s what I’ve achieved:

1️⃣ Enhanced Data Accuracy:

Curated a more focused, high-quality dataset

Applied rigorous data cleaning methods

Ensured a diverse range of prompt styles and topics

2️⃣ Retrained Llama-3.2 3B Model:

Fine-tuned with the improved dataset

Optimized training parameters for better performance

Achieved more nuanced and precise prompt generation

3️⃣ Custom ComfyUI Node:
Developed to enable seamless integration for text-to-image models.

🔗 Resources:

Fine-tuned models on Hugging Face:https://huggingface.co/ABDALLALSWAITI/llm_3_2_flux_prompt

Ollama Hub model: https://ollama.com/abedalswaity7/flux-prompt

GitHub code: https://github.com/al-swaiti/ComfyUI-OllamaGemini

This project aims to push the boundaries of AI-assisted creativity by generating more effective, nuanced prompts for models like Flux.

Would love to connect with fellow AI enthusiasts! Have you worked on similar projects? How do you see this tech evolving?
output.gif

AIArt #MachineLearning #NLP #ComputerVision #AIInnovation #TechDevelopment #AIArt #MachineLearning #NLP #ComputerVision #FluxDev #AI