t5x/examples/t5/mt5/large.gin

# mT5 Large model.

include 't5x/examples/t5/mt5/base.gin'  # imports vocab, optimizer and model.

# ------------------- Network specification overrides --------------------------
network.Transformer.config = @network.T5Config()
network.T5Config:
  emb_dim = 1024
  num_heads = 16
  num_encoder_layers = 24
  num_decoder_layers = 24
  head_dim = 64
  mlp_dim = 2816