
Artificial intelligence research laboratory DeepReinforce has launched Ornith-1.5, a series of open models that can be partially self-trained. During development, a mechanism was used that allowed the system to independently create training tasks and formulate testing methods. Image source: ornith.ai










