Hello folks! Yesterday, DeepSeek did a huge update to their R1 model, bringing its performance on par with OpenAI's o3, o4-mini-high and Google's Gemini 2.5 Pro. They called the model 'DeepSeek-R1-0528' (which was when the model finishe…

Hey folks! Text-to-Speech (TTS) models have been pretty popular recently and one way to customize it (e.g. cloning a voice), is by fine-tuning the model. There are other methods however you do training, if you want speaking speed, phrasing, vocal…