The Model Card for DeciCoder 1B provides details about a 1 billion parameter decoderonly code completion model developed by Deci. The model was trained on Python, Java, and JavaScript subsets of Starcoder Training Dataset and uses Grouped Query Attention with a context window of 2048 tokens. It was
DeciLM7B is a decoderonly text generation model with 7.04 billion parameters, released by Deci under the Apache 2.0 license. It is the topperforming 7B base language model on the Open LLM Leaderboard and uses variable GroupedQuery Attention (GQA) to achieve a superior balance between accuracy an
DeciDiffusion 1.0 is an 820 million parameter latent diffusion model designed for texttoimage conversion. Trained initially on the LAIONv2 dataset and finetuned on the LAIONART dataset, the model's training involved advanced techniques to improve speed, training performance, and achieve super
DeciLM7Binstruct is a model for shortform instruction following, built by LoRA finetuning on the SlimOrca dataset. It is a derivative of the recently released DeciLM7B language model, a pretrained, highefficiency generative text model with 7 billion parameters. DeciLM7Binstruct is one of th