# Instruction-Tuned Model

An intelligent vision-language model fine-tuned from HuggingFaceTB/SmolVLM-Instruct, optimized for training speed using Unsloth and TRL libraries

Llama-3.3-70B-Instruct-FP8-dynamic is an optimized large language model. By quantizing activations and weights to the FP8 data type, it reduces GPU memory requirements and improves computational throughput, supporting commercial and research use in multiple languages.

A multimodal large language model (MLLM) specifically designed for interpreting electrocardiogram (ECG) images, capable of handling various ECG-related tasks from diverse data sources.