Modification of RWKV V6 RNN mode to support batches of data
increasing size of the model increases performance as expected
at the moment only the mnist input dim of 28 works, I haven't debugged 1 yet
Training using PyTorch Lightning.
Modification of RWKV V6 RNN mode to support batches of data
increasing size of the model increases performance as expected
at the moment only the mnist input dim of 28 works, I haven't debugged 1 yet
Training using PyTorch Lightning.