Model record / Language Model
YunaGPT
A self-trained decoder language-model project covering architecture experiments, tokenizer work, data preparation, pretraining, instruction tuning, and alignment studies.
YunaGPT represents model-pipeline engineering rather than a downloaded model showcase.
The full record will document tokenizer design, architecture choices, dataset work, pretraining, supervised fine-tuning, conversation training, preference experiments, failures, and lessons learned.
Related / Model registry
Continue through the training records
Related records are selected from shared model type and tag metadata.
model-diffusion-v1research
ModelDiffusion ModelSelf-Trained Diffusion Study
A diffusion-model research track focused on dataset work, training configuration, checkpoint progression, generated outputs, and evaluation.