Upload 2 files

clevr_data2023-03-02T11-07-16-lightning.yaml

@@ -0,0 +1,32 @@
+lightning:
+  find_unused_parameters: false
+  modelcheckpoint:
+    params:
+      monitor: train/loss_simple
+      save_top_k: -1
+      every_n_train_steps: 405
+  callbacks:
+    image_logger:
+      target: main.ImageLogger
+      params:
+        batch_frequency: 10000000
+        log_on_batch_idx: true
+        max_images: 4
+        increase_log_steps: false
+        log_images_kwargs:
+          inpaint: false
+          plot_denoise_rows: false
+          plot_progressive_rows: false
+          plot_diffusion_rows: false
+          'N': 4
+          ddim_steps: 50
+  trainer:
+    benchmark: true
+    max_steps: 20000
+    check_val_every_n_epoch: 1000
+    num_sanity_val_steps: 0
+    accumulate_grad_batches: 8
+    accelerator: gpu
+    strategy: ddp
+    num_nodes: 2
+    gpus: 0,1,2,3,4,5,6,7

clevr_data2023-03-02T11-07-16-project.yaml

@@ -0,0 +1,116 @@
+model:
+  base_learning_rate: 0.0001
+  target: ldm.models.diffusion.ddpm.LatentInpaintDiffusion
+  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
+    conditioning_key: hybrid
+    monitor: val/loss_simple_ema
+    scale_factor: 0.18215
+    finetune_keys: null
+    scheduler_config:
+      target: ldm.lr_scheduler.LambdaLinearScheduler
+      params:
+        warm_up_steps:
+        - 1
+        cycle_lengths:
+        - 10000000000000
+        f_start:
+        - 1.0e-06
+        f_max:
+        - 1.0
+        f_min:
+        - 1.0
+    unet_config:
+      target: ldm.modules.diffusionmodules.openaimodel.UNetModel
+      params:
+        image_size: 32
+        in_channels: 9
+        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: true
+        legacy: false
+    first_stage_config:
+      target: ldm.models.autoencoder.AutoencoderKL
+      params:
+        embed_dim: 4
+        monitor: val/rec_loss
+        ddconfig:
+          double_z: true
+          z_channels: 4
+          resolution: 512
+          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: ldm.modules.encoders.modules.FrozenCLIPEmbedder
+      params:
+        with_bbox: true
+        num_bins: 1000
+        extend_outputlen: 616
+        max_length: 616
+    ckpt_path: preload_model_checkpoints/SD_checkpoints/sd-v1-5-inpainting.ckpt
+data:
+  target: main.DataModuleFromConfig
+  params:
+    batch_size: 1
+    num_workers: 8
+    wrap: false
+    train:
+      target: ldm.data.clevr.CLEVRDataset
+      params:
+        size: 512
+        old_clevr: true
+        set: train
+        with_bbox: true
+        num_bins: 1000
+        flip_p: 0.5
+        box_descp: caption
+        max_src_length: 616
+        has_global_caption: false
+        inpaint: iterative
+    validation:
+      target: ldm.data.clevr.CLEVRDataset
+      params:
+        size: 512
+        old_clevr: true
+        set: test
+        with_bbox: true
+        num_bins: 1000
+        flip_p: 0.5
+        box_descp: caption
+        max_src_length: 616
+        has_global_caption: false
+        inpaint: iterative