This is a fine-tuned version of Meta-Llama-3.1-8B, trained using Unsloth and merged into a single GGUF (Q8_0) file for efficient local inference.

Repo: talspec2/red-string

Model Details

Base Model: Meta Llama 3.1 8B Instruct
Quantization: Q8_0 (8-bit, medium compressed)
Format: GGUF (Ready for llama.cpp, LM Studio, Ollama)
Fine-Tuning: SFT (Supervised Fine-Tuning) with PEFT/LoRA

Prompt Template

You must use this structure for the best results:

text

### Instruction:
{INSTRUCTION}

### Input:
{INPUT}

### Response:
{LEAVE BLANK}

tspec2/redstring-8gb

Red String (Llama 3.1 8B Fine-Tuned)

Model Details

Prompt Template

tspec2/
redstring-8gb