# English Task Optimization

Llama 3.2 11B vision instruction fine-tuned model optimized with Unsloth, using 4-bit quantization technology, achieving 2x faster training speed

A fine-tuned version of the Qwen2-VL-7B model, trained using Unsloth and Huggingface's TRL library, achieving a 2x speed improvement.

This is a GGUF format model converted from MBZUAI/LaMini-Flan-T5-248M, suitable for text generation tasks.

WebLINX is a model based on Llama3, focusing on agent tasks and language understanding, suitable for English environments.