Another announcement at AMD today beyond the open-source Linux driver fun for the Radeon RX 9070 series is announcing the open-sourcing of Instella as their new fully open 3B parameter language models.
AMD Instella represents “fully open state-of-the-art 3-billion-parameter language models (LMs).” These models were trained on AMD Instinct MI300X GPUs and according to AMD’s published data delivers competitive performance to the likes of Llama 3.2 3B, Gemma-2 2B, and Qwen 2.5 3B.
The AMD Instella models were trained from scratch on Instinct MI300X hardware and are fully open-source:
“Fully open and accessible: Fully open-source release of model weights, training hyperparameters, datasets, and code, fostering innovation and collaboration within the AI community.
…
By fully open sourcing the Instella models, including weights, training configurations, datasets, and code, we aim to foster innovation and collaboration within the AI community. We believe that transparency, reproducibility and accessibility are key drivers of progress in AI research and development. We invite developers, researchers, and AI enthusiasts to explore Instella, contribute to its ongoing improvement, and join us in pushing the boundaries of what is possible with language models.”
Those wanting to learn more about the AMD Instella language models can do so via the rocm.blogs.amd.com. AMD Instella is hosted on GitHub.
