| model: | |
| transformer_config: | |
| checkpoint_path: ./checkpoints/kny_video_full_gpt2_medium/checkpoint | |
| remaining_frames_method: "own_embeddings" | |
| transformer_type: "gpt2-medium" | |
| first_stage_config: | |
| checkpoint_path: ./checkpoints/kny_image_full_vgg19/checkpoint | |
| vqvae_config: | |
| beta: 0.25 | |
| num_embeddings: 50257 | |
| embedding_dim: 128 | |
| autoencoder_config: | |
| z_channels: 512 | |
| channels: 32 | |
| channels_multiplier: | |
| - 2 | |
| - 4 | |
| - 8 | |
| - 8 | |
| num_res_blocks: 1 | |
| attention_resolution: | |
| - 16 | |
| resolution: 128 | |
| dropout: 0.0 | |
| discriminator_config: | |
| num_layers: 3 | |
| filters: 64 | |
| loss_config: | |
| discriminator: | |
| loss: "hinge" | |
| factor: 1.0 | |
| iter_start: 16200 | |
| weight: 0.3 | |
| vqvae: | |
| codebook_weight: 1.0 | |
| perceptual_weight: 4.0 | |
| perceptual_loss: "vgg19" | |
| train: | |
| batch_size: 64 | |
| accumulation_size: 1 | |
| n_epochs: 500 | |
| len_x_train: 28213 | |
| warmup_epoch_percentage: 0.15 | |
| lr_start: 5e-6 | |
| lr_max: 1e-4 | |
| perceptual_loss_weight: 1.0 | |
| n_frames_before: 5 | |
| stop_ground_truth_after_epoch: 200 | |