| model: | |
| base_learning_rate: 1.0e-05 | |
| target: src.imagen_hub.pipelines.custom_diffusion.custom_diffusion_src.src.model.CustomDiffusion | |
| params: | |
| linear_start: 0.00085 | |
| linear_end: 0.012 | |
| num_timesteps_cond: 1 | |
| log_every_t: 200 | |
| timesteps: 1000 | |
| first_stage_key: image | |
| cond_stage_key: caption | |
| image_size: 64 | |
| channels: 4 | |
| cond_stage_trainable: true | |
| add_token: true | |
| freeze_model: crossattn-kv | |
| conditioning_key: crossattn | |
| monitor: val/loss_simple_ema | |
| scale_factor: 0.18215 | |
| use_ema: false | |
| unet_config: | |
| target: src.imagen_hub.pipelines.custom_diffusion.custom_diffusion_src.stable-diffusion.ldm.modules.diffusionmodules.openaimodel.UNetModel | |
| params: | |
| image_size: 64 | |
| in_channels: 4 | |
| out_channels: 4 | |
| model_channels: 320 | |
| attention_resolutions: | |
| - 4 | |
| - 2 | |
| - 1 | |
| num_res_blocks: 2 | |
| channel_mult: | |
| - 1 | |
| - 2 | |
| - 4 | |
| - 4 | |
| num_heads: 8 | |
| use_spatial_transformer: true | |
| transformer_depth: 1 | |
| context_dim: 768 | |
| use_checkpoint: false | |
| legacy: false | |
| first_stage_config: | |
| target: src.imagen_hub.pipelines.custom_diffusion.custom_diffusion_src.stable-diffusion.ldm.models.autoencoder.AutoencoderKL | |
| params: | |
| embed_dim: 4 | |
| monitor: val/rec_loss | |
| ddconfig: | |
| double_z: true | |
| z_channels: 4 | |
| resolution: 256 | |
| in_channels: 3 | |
| out_ch: 3 | |
| ch: 128 | |
| ch_mult: | |
| - 1 | |
| - 2 | |
| - 4 | |
| - 4 | |
| num_res_blocks: 2 | |
| attn_resolutions: [] | |
| dropout: 0.0 | |
| lossconfig: | |
| target: torch.nn.Identity | |
| cond_stage_config: | |
| target: src.imagen_hub.pipelines.custom_diffusion.custom_diffusion_src.src.custom_modules.FrozenCLIPEmbedderWrapper | |
| params: | |
| modifier_token: <new1>+<new2> | |
| ckpt_path: null | |
| data: | |
| target: train.DataModuleFromConfig | |
| params: | |
| batch_size: 2 | |
| num_workers: 4 | |
| wrap: false | |
| train: | |
| target: src.finetune_data.MaskBase | |
| params: | |
| size: 512 | |
| caption: <new1> dog | |
| reg_caption: real_reg/samples_dog/caption.txt | |
| datapath: data/dog | |
| reg_datapath: real_reg/samples_dog/images.txt | |
| train2: | |
| target: src.finetune_data.MaskBase | |
| params: | |
| size: 512 | |
| caption: <new2> car | |
| reg_caption: real_reg/samples_car/caption.txt | |
| datapath: data/car | |
| reg_datapath: real_reg/samples_car/images.txt | |
| validation: | |
| target: src.finetune_data.MaskBase | |
| params: | |
| size: 512 | |
| caption: <new1> dog | |
| reg_caption: real_reg/samples_dog/caption.txt | |
| datapath: data/dog | |
| reg_datapath: real_reg/samples_dog/images.txt | |