Text Generation
Safetensors
Spanish
cybersecurity
seguridad-informatica
qwen2.5
lora
sft
unsloth
conversational
Instructions to use murdok1982/MurdokLLmHack-LoRA with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Local Apps
- Unsloth Studio new
How to use murdok1982/MurdokLLmHack-LoRA with Unsloth Studio:
Install Unsloth Studio (macOS, Linux, WSL)
curl -fsSL https://unsloth.ai/install.sh | sh # Run unsloth studio unsloth studio -H 0.0.0.0 -p 8888 # Then open http://localhost:8888 in your browser # Search for murdok1982/MurdokLLmHack-LoRA to start chatting
Install Unsloth Studio (Windows)
irm https://unsloth.ai/install.ps1 | iex # Run unsloth studio unsloth studio -H 0.0.0.0 -p 8888 # Then open http://localhost:8888 in your browser # Search for murdok1982/MurdokLLmHack-LoRA to start chatting
Using HuggingFace Spaces for Unsloth
# No setup required # Open https://huggingface.co/spaces/unsloth/studio in your browser # Search for murdok1982/MurdokLLmHack-LoRA to start chatting
Load model with FastModel
pip install unsloth from unsloth import FastModel model, tokenizer = FastModel.from_pretrained( model_name="murdok1982/MurdokLLmHack-LoRA", max_seq_length=2048, )
| license: mit | |
| language: | |
| - es | |
| tags: | |
| - cybersecurity | |
| - seguridad-informatica | |
| - qwen2.5 | |
| - lora | |
| - sft | |
| - unsloth | |
| base_model: unsloth/qwen2.5-1.5b-bnb-4bit | |
| pipeline_tag: text-generation | |
| # MurdokLLmHack — Modelo de Ciberseguridad Fine-Tuned | |
| Modelo fine-tuned sobre Qwen2.5-1.5B con +16,000 pares Q&A extraídos de 59 documentos técnicos de ciberseguridad. | |
| ## Uso con Ollama (recomendado) | |
| ```bash | |
| ollama run murdokllmhack | |
| ``` | |
| ## Uso con Transformers + PEFT | |
| ```python | |
| from transformers import AutoModelForCausalLM, AutoTokenizer | |
| from peft import PeftModel | |
| import torch | |
| base = AutoModelForCausalLM.from_pretrained( | |
| 'unsloth/qwen2.5-1.5b-bnb-4bit', | |
| device_map='auto', | |
| torch_dtype=torch.float16 | |
| ) | |
| tokenizer = AutoTokenizer.from_pretrained('unsloth/qwen2.5-1.5b-bnb-4bit') | |
| model = PeftModel.from_pretrained(base, 'murdok1982/MurdokLLmHack-LoRA') | |
| prompt = '<|im_start|>system\nEres un experto en ciberseguridad.<|im_end|>\n<|im_start|>user\nQue es un firewall?<|im_end|>\n<|im_start|>assistant' | |
| inputs = tokenizer(prompt, return_tensors='pt').to('cuda') | |
| outputs = model.generate(**inputs, max_new_tokens=256, temperature=0.7) | |
| print(tokenizer.decode(outputs[0], skip_special_tokens=True)) | |
| ``` | |
| ## Uso con GGUF (inferencia local CPU) | |
| Descarga el GGUF y usa: | |
| ```bash | |
| ollama create murdokllmhack -f Modelfile | |
| ollama run murdokllmhack | |
| ``` | |
| ## Detalles del Fine-Tuning | |
| - **Base:** Qwen2.5-1.5B | |
| - **Dataset:** 16,026 train / 1,781 validation (formato ChatML) | |
| - **Entrenamiento:** Unsloth + LoRA (r=32), 3 epochs, T4 Colab | |
| - **Cuantizacion:** Q8_0 (GGUF), fp16 (merge) | |
| - **Contexto:** 131,072 tokens | |
| ## Contacto | |
| - **Email:** gustavolobatoclara@gmail.com | |
| - **LinkedIn:** https://www.linkedin.com/in/gustavo-lobato-clara1982/ | |
| - **Dataset:** https://huggingface.co/datasets/murdok1982/formacion-seguridad-qa | |