AMD

A 16B-A3B MoE trained by AMD on Instinct cards. AMD also provides all the different stages, from the base to the SFT checkpoints, as well as MidTrain and DPO.

A fully open 3B model trained on the OLMo recipe and tools. The model was trained on the MI300X, AMD's equivalent to the H100. Their evaluations place it as competitive with some other small instruct models (Gemma 2 and Llama 3.2), which we'll follow in community adoption.

Using our OLMo code with the accompanying dataset and our Tülu data for SFT, AMD has released a 1B model to showcase the training capabilities of their GPUs.