transformer.yaml

model:
  base_learning_rate: 4.5e-06
  target: taming_comb.models.cond_transformer.Net2NetTransformer
  params:
    cond_stage_key: coord
    transformer_config:
      target: taming_comb.modules.transformer.mingpt.GPT
      params:
        vocab_size: 512
        block_size: 512
        n_layer: 24
        n_head: 16
        n_embd: 512

    first_stage_model_config:
      target: taming_comb.models.vqgan.VQModelCrossGAN_ADAIN
      params:
        embed_dim: 256
        n_embed: 256
        ddconfig:
          double_z: False
          z_channels: 256
          resolution: 256
          in_channels: 3
          out_ch: 3
          ch: 128
          ch_mult: [ 1,1,2,2,4]  # num_down = len(ch_mult)-1
          num_res_blocks: 2
          attn_resolutions: [16]
          dropout: 0.0
        lossconfig:
          target: taming_comb.modules.losses.vqperceptual.VQLPIPSWithDiscriminator
          params:
            disc_conditional: False
            disc_in_channels: 3
            disc_start: 0 #301
            disc_factor: 1.0 #0.8 / 0.3
            disc_weight: 0.1 # 0.5 / 0.8
            codebook_weight: 1.0

    cond_stage_config:
      target: taming_comb.modules.misc.coord.CoordStage
      params:
        n_embed: 512
        down_factor: 16

    f_path: None
    device: '0'