Upload 40 files

DepthAnything_vitb.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:851ed48c54d3bc91be762169662ae6df3b522d741a326d8d0cfad90deb0dd893
+size 391269536

DepthAnything_vitl.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:09e3c568c99c6437a53ccdbb59b3d2ec7a918ef5f74a775dfca6a14ee3ab3f57
+size 1343379696

DepthAnything_vits.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ecc1f60f6e66ed39c7a3bb6395f1e5fd649f95505833d82903ce79df4eb1a42a
+size 100247904

ZoeDepthv1.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9b6f2a136ea7dfe1368638b8cdee5d01b687158d9e08e7c4ee0bc21613f94dc4
+size 1443428517

depthanything_vitb_u4k/coarse_pretrain/20240315_095516.log

@@ -0,0 +1,1024 @@
+2024/03/15 09:55:26 - patchstitcher - INFO - 
+------------------------------------------------------------
+System environment:
+    sys.platform: linux
+    Python: 3.8.18 | packaged by conda-forge | (default, Oct 10 2023, 15:44:36) [GCC 12.3.0]
+    CUDA available: True
+    numpy_random_seed: 621
+    GPU 0,1,2,3: NVIDIA A100-SXM4-80GB
+    CUDA_HOME: /sw/rl9g/cuda/11.8/rl9_binary
+    NVCC: Cuda compilation tools, release 11.8, V11.8.89
+    GCC: gcc (GCC) 11.3.1 20220421 (Red Hat 11.3.1-2)
+    PyTorch: 2.1.2
+    PyTorch compiling details: PyTorch built with:
+  - GCC 9.3
+  - C++ Version: 201703
+  - Intel(R) oneAPI Math Kernel Library Version 2022.1-Product Build 20220311 for Intel(R) 64 architecture applications
+  - Intel(R) MKL-DNN v3.1.1 (Git Hash 64f6bcbcbab628e96f33a62c3e975f8535a7bde4)
+  - OpenMP 201511 (a.k.a. OpenMP 4.5)
+  - LAPACK is enabled (usually provided by MKL)
+  - NNPACK is enabled
+  - CPU capability usage: AVX2
+  - CUDA Runtime 11.8
+  - NVCC architecture flags: -gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_61,code=sm_61;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_80,code=sm_80;-gencode;arch=compute_86,code=sm_86;-gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_90,code=sm_90;-gencode;arch=compute_37,code=compute_37
+  - CuDNN 8.7
+  - Magma 2.6.1
+  - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=11.8, CUDNN_VERSION=8.7.0, CXX_COMPILER=/opt/rh/devtoolset-9/root/usr/bin/c++, CXX_FLAGS= -D_GLIBCXX_USE_CXX11_ABI=0 -fabi-version=11 -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOROCTRACER -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -O2 -fPIC -Wall -Wextra -Werror=return-type -Werror=non-virtual-dtor -Werror=bool-operation -Wnarrowing -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wno-unused-parameter -Wno-unused-function -Wno-unused-result -Wno-strict-overflow -Wno-strict-aliasing -Wno-stringop-overflow -Wno-psabi -Wno-error=pedantic -Wno-error=old-style-cast -Wno-invalid-partial-specialization -Wno-unused-private-field -Wno-aligned-allocation-unavailable -Wno-missing-braces -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Werror=cast-function-type -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_DISABLE_GPU_ASSERTS=ON, TORCH_VERSION=2.1.2, USE_CUDA=ON, USE_CUDNN=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=ON, USE_NNPACK=ON, USE_OPENMP=ON, USE_ROCM=OFF, 
+
+    TorchVision: 0.16.2
+    OpenCV: 4.8.1
+    MMEngine: 0.10.2
+
+Runtime environment:
+    cudnn_benchmark: True
+    mp_cfg: {'mp_start_method': 'forkserver'}
+    dist_cfg: {'backend': 'nccl'}
+    seed: 621
+    Distributed launcher: pytorch
+    Distributed training: True
+    GPU number: 4
+------------------------------------------------------------
+
+2024/03/15 09:55:26 - patchstitcher - INFO - Config:
+collect_input_args = [
+    'image_lr',
+    'crops_image_hr',
+    'depth_gt',
+    'crop_depths',
+    'bboxs',
+    'image_hr',
+]
+convert_syncbn = True
+debug = False
+env_cfg = dict(
+    cudnn_benchmark=True,
+    dist_cfg=dict(backend='nccl'),
+    mp_cfg=dict(mp_start_method='forkserver'))
+find_unused_parameters = True
+general_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        dataset_name='', gt_dir=None, rgb_image_dir='', type='ImageDataset'),
+    num_workers=2)
+launcher = 'pytorch'
+log_name = 'coarse_pretrain'
+max_depth = 80
+min_depth = 0.001
+model = dict(
+    coarse_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        depth_anything=True,
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            392,
+            518,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='vitb',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource='local::./work_dir/DepthAnything_vitb.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='DA-ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    fine_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        depth_anything=True,
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            392,
+            518,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='vitb',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource='local::./work_dir/DepthAnything_vitb.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='DA-ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    max_depth=80,
+    min_depth=0.001,
+    sigloss=dict(type='SILogLoss'),
+    target='coarse',
+    type='BaselinePretrain')
+optim_wrapper = dict(
+    clip_grad=dict(max_norm=0.1, norm_type=2, type='norm'),
+    optimizer=dict(lr=4e-06, type='AdamW', weight_decay=0.01),
+    paramwise_cfg=dict(bypass_duplicate=True, custom_keys=dict()))
+param_scheduler = dict(
+    base_momentum=0.85,
+    cycle_momentum=True,
+    div_factor=1,
+    final_div_factor=10000,
+    max_momentum=0.95,
+    pct_start=0.5,
+    three_phase=False)
+project = 'patchfusion'
+tags = [
+    'coarse',
+    'da',
+    'vitb',
+]
+test_in_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+test_out_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test_out.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+train_cfg = dict(
+    eval_start=0,
+    log_interval=100,
+    max_epochs=24,
+    save_checkpoint_interval=24,
+    train_log_img_interval=500,
+    val_interval=2,
+    val_log_img_interval=50,
+    val_type='epoch_base')
+train_dataloader = dict(
+    batch_size=4,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='train',
+        resize_mode='depth-anything',
+        split='./data/u4k/splits/train.txt',
+        transform_cfg=dict(
+            degree=1.0, network_process_size=[
+                392,
+                518,
+            ], random_crop=True),
+        type='UnrealStereo4kDataset'),
+    num_workers=4)
+val_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        resize_mode='depth-anything',
+        split='./data/u4k/splits/val.txt',
+        transform_cfg=dict(degree=1.0, network_process_size=[
+            392,
+            518,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+work_dir = './work_dir/depthanything_vitb_u4k/coarse_pretrain'
+zoe_depth_config = dict(
+    attractor_alpha=1000,
+    attractor_gamma=2,
+    attractor_kind='mean',
+    attractor_type='inv',
+    aug=True,
+    bin_centers_type='softplus',
+    bin_embedding_dim=128,
+    clip_grad=0.1,
+    dataset='nyu',
+    depth_anything=True,
+    distributed=True,
+    do_resize=False,
+    force_keep_ar=True,
+    freeze_midas_bn=True,
+    gpu='NULL',
+    img_size=[
+        392,
+        518,
+    ],
+    inverse_midas=False,
+    log_images_every=0.1,
+    max_depth=80,
+    max_temp=50.0,
+    max_translation=100,
+    memory_efficient=True,
+    midas_model_type='vitb',
+    min_depth=0.001,
+    min_temp=0.0212,
+    model='zoedepth',
+    n_attractors=[
+        16,
+        8,
+        4,
+        1,
+    ],
+    n_bins=64,
+    name='ZoeDepth',
+    notes='',
+    output_distribution='logbinomial',
+    prefetch=False,
+    pretrained_resource='local::./work_dir/DepthAnything_vitb.pt',
+    print_losses=False,
+    project='ZoeDepth',
+    random_crop=False,
+    random_translate=False,
+    root='.',
+    save_dir='',
+    shared_dict='NULL',
+    tags='',
+    train_midas=True,
+    translate_prob=0.2,
+    type='DA-ZoeDepth',
+    uid='NULL',
+    use_amp=False,
+    use_pretrained_midas=True,
+    use_shared_dict=False,
+    validate_every=0.25,
+    version_name='v1',
+    workers=16)
+
+2024/03/15 09:55:28 - patchstitcher - INFO - Loading deepnet from local::./work_dir/DepthAnything_vitb.pt
+2024/03/15 09:55:28 - patchstitcher - INFO - Current zoedepth.core.prep.resizer is <class 'torch.nn.modules.linear.Identity'>
+2024/03/15 09:55:28 - patchstitcher - INFO - DistributedDataParallel(
+  (module): BaselinePretrain(
+    (coarse_branch): ZoeDepth(
+      (core): DepthAnythingCore(
+        (core): DPT_DINOv2(
+          (pretrained): DinoVisionTransformer(
+            (patch_embed): PatchEmbed(
+              (proj): Conv2d(3, 768, kernel_size=(14, 14), stride=(14, 14))
+              (norm): Identity()
+            )
+            (blocks): ModuleList(
+              (0-11): 12 x NestedTensorBlock(
+                (norm1): LayerNorm((768,), eps=1e-06, elementwise_affine=True)
+                (attn): MemEffAttention(
+                  (qkv): Linear(in_features=768, out_features=2304, bias=True)
+                  (attn_drop): Dropout(p=0.0, inplace=False)
+                  (proj): Linear(in_features=768, out_features=768, bias=True)
+                  (proj_drop): Dropout(p=0.0, inplace=False)
+                )
+                (ls1): LayerScale()
+                (drop_path1): Identity()
+                (norm2): LayerNorm((768,), eps=1e-06, elementwise_affine=True)
+                (mlp): Mlp(
+                  (fc1): Linear(in_features=768, out_features=3072, bias=True)
+                  (act): GELU(approximate='none')
+                  (fc2): Linear(in_features=3072, out_features=768, bias=True)
+                  (drop): Dropout(p=0.0, inplace=False)
+                )
+                (ls2): LayerScale()
+                (drop_path2): Identity()
+              )
+            )
+            (norm): LayerNorm((768,), eps=1e-06, elementwise_affine=True)
+            (head): Identity()
+          )
+          (depth_head): DPTHead(
+            (projects): ModuleList(
+              (0): Conv2d(768, 96, kernel_size=(1, 1), stride=(1, 1))
+              (1): Conv2d(768, 192, kernel_size=(1, 1), stride=(1, 1))
+              (2): Conv2d(768, 384, kernel_size=(1, 1), stride=(1, 1))
+              (3): Conv2d(768, 768, kernel_size=(1, 1), stride=(1, 1))
+            )
+            (resize_layers): ModuleList(
+              (0): ConvTranspose2d(96, 96, kernel_size=(4, 4), stride=(4, 4))
+              (1): ConvTranspose2d(192, 192, kernel_size=(2, 2), stride=(2, 2))
+              (2): Identity()
+              (3): Conv2d(768, 768, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1))
+            )
+            (scratch): Module(
+              (layer1_rn): Conv2d(96, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
+              (layer2_rn): Conv2d(192, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
+              (layer3_rn): Conv2d(384, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
+              (layer4_rn): Conv2d(768, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
+              (refinenet1): FeatureFusionBlock(
+                (out_conv): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+                (resConfUnit1): ResidualConvUnit(
+                  (conv1): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (resConfUnit2): ResidualConvUnit(
+                  (conv1): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (skip_add): FloatFunctional(
+                  (activation_post_process): Identity()
+                )
+              )
+              (refinenet2): FeatureFusionBlock(
+                (out_conv): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+                (resConfUnit1): ResidualConvUnit(
+                  (conv1): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (resConfUnit2): ResidualConvUnit(
+                  (conv1): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (skip_add): FloatFunctional(
+                  (activation_post_process): Identity()
+                )
+              )
+              (refinenet3): FeatureFusionBlock(
+                (out_conv): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+                (resConfUnit1): ResidualConvUnit(
+                  (conv1): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (resConfUnit2): ResidualConvUnit(
+                  (conv1): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (skip_add): FloatFunctional(
+                  (activation_post_process): Identity()
+                )
+              )
+              (refinenet4): FeatureFusionBlock(
+                (out_conv): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+                (resConfUnit1): ResidualConvUnit(
+                  (conv1): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (resConfUnit2): ResidualConvUnit(
+                  (conv1): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (skip_add): FloatFunctional(
+                  (activation_post_process): Identity()
+                )
+              )
+              (output_conv1): Conv2d(128, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+              (output_conv2): Sequential(
+                (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                (1): ReLU(inplace=True)
+                (2): Conv2d(32, 1, kernel_size=(1, 1), stride=(1, 1))
+                (3): ReLU(inplace=True)
+                (4): Identity()
+              )
+            )
+          )
+        )
+      )
+      (conv2): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+      (seed_bin_regressor): SeedBinRegressorUnnormed(
+        (_net): Sequential(
+          (0): Conv2d(128, 256, kernel_size=(1, 1), stride=(1, 1))
+          (1): ReLU(inplace=True)
+          (2): Conv2d(256, 64, kernel_size=(1, 1), stride=(1, 1))
+          (3): Softplus(beta=1, threshold=20)
+        )
+      )
+      (seed_projector): Projector(
+        (_net): Sequential(
+          (0): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+          (1): ReLU(inplace=True)
+          (2): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+        )
+      )
+      (projectors): ModuleList(
+        (0-3): 4 x Projector(
+          (_net): Sequential(
+            (0): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+            (1): ReLU(inplace=True)
+            (2): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+          )
+        )
+      )
+      (attractors): ModuleList(
+        (0): AttractorLayerUnnormed(
+          (_net): Sequential(
+            (0): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+            (1): ReLU(inplace=True)
+            (2): Conv2d(128, 16, kernel_size=(1, 1), stride=(1, 1))
+            (3): Softplus(beta=1, threshold=20)
+          )
+        )
+        (1): AttractorLayerUnnormed(
+          (_net): Sequential(
+            (0): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+            (1): ReLU(inplace=True)
+            (2): Conv2d(128, 8, kernel_size=(1, 1), stride=(1, 1))
+            (3): Softplus(beta=1, threshold=20)
+          )
+        )
+        (2): AttractorLayerUnnormed(
+          (_net): Sequential(
+            (0): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+            (1): ReLU(inplace=True)
+            (2): Conv2d(128, 4, kernel_size=(1, 1), stride=(1, 1))
+            (3): Softplus(beta=1, threshold=20)
+          )
+        )
+        (3): AttractorLayerUnnormed(
+          (_net): Sequential(
+            (0): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+            (1): ReLU(inplace=True)
+            (2): Conv2d(128, 1, kernel_size=(1, 1), stride=(1, 1))
+            (3): Softplus(beta=1, threshold=20)
+          )
+        )
+      )
+      (conditional_log_binomial): ConditionalLogBinomial(
+        (log_binomial_transform): LogBinomial()
+        (mlp): Sequential(
+          (0): Conv2d(161, 80, kernel_size=(1, 1), stride=(1, 1))
+          (1): GELU(approximate='none')
+          (2): Conv2d(80, 4, kernel_size=(1, 1), stride=(1, 1))
+          (3): Softplus(beta=1, threshold=20)
+        )
+      )
+    )
+    (sigloss): SILogLoss()
+  )
+)
+2024/03/15 09:55:34 - patchstitcher - INFO - successfully init trainer
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.cls_token
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.pos_embed
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.mask_token
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.patch_embed.proj.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.patch_embed.proj.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.0.norm1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.0.norm1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.0.attn.qkv.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.0.attn.qkv.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.0.attn.proj.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.0.attn.proj.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.0.ls1.gamma
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.0.norm2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.0.norm2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.0.mlp.fc1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.0.mlp.fc1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.0.mlp.fc2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.0.mlp.fc2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.0.ls2.gamma
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.1.norm1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.1.norm1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.1.attn.qkv.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.1.attn.qkv.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.1.attn.proj.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.1.attn.proj.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.1.ls1.gamma
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.1.norm2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.1.norm2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.1.mlp.fc1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.1.mlp.fc1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.1.mlp.fc2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.1.mlp.fc2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.1.ls2.gamma
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.2.norm1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.2.norm1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.2.attn.qkv.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.2.attn.qkv.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.2.attn.proj.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.2.attn.proj.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.2.ls1.gamma
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.2.norm2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.2.norm2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.2.mlp.fc1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.2.mlp.fc1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.2.mlp.fc2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.2.mlp.fc2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.2.ls2.gamma
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.3.norm1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.3.norm1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.3.attn.qkv.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.3.attn.qkv.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.3.attn.proj.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.3.attn.proj.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.3.ls1.gamma
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.3.norm2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.3.norm2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.3.mlp.fc1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.3.mlp.fc1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.3.mlp.fc2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.3.mlp.fc2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.3.ls2.gamma
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.4.norm1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.4.norm1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.4.attn.qkv.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.4.attn.qkv.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.4.attn.proj.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.4.attn.proj.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.4.ls1.gamma
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.4.norm2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.4.norm2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.4.mlp.fc1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.4.mlp.fc1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.4.mlp.fc2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.4.mlp.fc2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.4.ls2.gamma
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.5.norm1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.5.norm1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.5.attn.qkv.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.5.attn.qkv.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.5.attn.proj.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.5.attn.proj.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.5.ls1.gamma
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.5.norm2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.5.norm2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.5.mlp.fc1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.5.mlp.fc1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.5.mlp.fc2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.5.mlp.fc2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.5.ls2.gamma
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.6.norm1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.6.norm1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.6.attn.qkv.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.6.attn.qkv.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.6.attn.proj.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.6.attn.proj.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.6.ls1.gamma
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.6.norm2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.6.norm2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.6.mlp.fc1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.6.mlp.fc1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.6.mlp.fc2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.6.mlp.fc2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.6.ls2.gamma
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.7.norm1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.7.norm1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.7.attn.qkv.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.7.attn.qkv.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.7.attn.proj.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.7.attn.proj.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.7.ls1.gamma
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.7.norm2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.7.norm2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.7.mlp.fc1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.7.mlp.fc1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.7.mlp.fc2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.7.mlp.fc2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.7.ls2.gamma
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.8.norm1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.8.norm1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.8.attn.qkv.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.8.attn.qkv.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.8.attn.proj.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.8.attn.proj.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.8.ls1.gamma
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.8.norm2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.8.norm2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.8.mlp.fc1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.8.mlp.fc1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.8.mlp.fc2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.8.mlp.fc2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.8.ls2.gamma
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.9.norm1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.9.norm1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.9.attn.qkv.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.9.attn.qkv.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.9.attn.proj.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.9.attn.proj.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.9.ls1.gamma
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.9.norm2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.9.norm2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.9.mlp.fc1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.9.mlp.fc1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.9.mlp.fc2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.9.mlp.fc2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.9.ls2.gamma
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.10.norm1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.10.norm1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.10.attn.qkv.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.10.attn.qkv.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.10.attn.proj.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.10.attn.proj.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.10.ls1.gamma
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.10.norm2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.10.norm2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.10.mlp.fc1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.10.mlp.fc1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.10.mlp.fc2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.10.mlp.fc2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.10.ls2.gamma
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.11.norm1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.11.norm1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.11.attn.qkv.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.11.attn.qkv.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.11.attn.proj.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.11.attn.proj.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.11.ls1.gamma
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.11.norm2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.11.norm2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.11.mlp.fc1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.11.mlp.fc1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.11.mlp.fc2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.11.mlp.fc2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.11.ls2.gamma
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.norm.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.norm.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.projects.0.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.projects.0.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.projects.1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.projects.1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.projects.2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.projects.2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.projects.3.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.projects.3.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.resize_layers.0.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.resize_layers.0.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.resize_layers.1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.resize_layers.1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.resize_layers.3.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.resize_layers.3.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.layer1_rn.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.layer2_rn.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.layer3_rn.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.layer4_rn.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet1.out_conv.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet1.out_conv.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet1.resConfUnit1.conv1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet1.resConfUnit1.conv1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet1.resConfUnit1.conv2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet1.resConfUnit1.conv2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet1.resConfUnit2.conv1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet1.resConfUnit2.conv1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet1.resConfUnit2.conv2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet1.resConfUnit2.conv2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet2.out_conv.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet2.out_conv.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet2.resConfUnit1.conv1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet2.resConfUnit1.conv1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet2.resConfUnit1.conv2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet2.resConfUnit1.conv2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet2.resConfUnit2.conv1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet2.resConfUnit2.conv1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet2.resConfUnit2.conv2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet2.resConfUnit2.conv2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet3.out_conv.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet3.out_conv.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet3.resConfUnit1.conv1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet3.resConfUnit1.conv1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet3.resConfUnit1.conv2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet3.resConfUnit1.conv2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet3.resConfUnit2.conv1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet3.resConfUnit2.conv1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet3.resConfUnit2.conv2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet3.resConfUnit2.conv2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet4.out_conv.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet4.out_conv.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet4.resConfUnit1.conv1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet4.resConfUnit1.conv1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet4.resConfUnit1.conv2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet4.resConfUnit1.conv2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet4.resConfUnit2.conv1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet4.resConfUnit2.conv1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet4.resConfUnit2.conv2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet4.resConfUnit2.conv2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.output_conv1.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.output_conv1.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.output_conv2.0.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.output_conv2.0.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.output_conv2.2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.output_conv2.2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.conv2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.conv2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.seed_bin_regressor._net.0.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.seed_bin_regressor._net.0.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.seed_bin_regressor._net.2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.seed_bin_regressor._net.2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.seed_projector._net.0.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.seed_projector._net.0.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.seed_projector._net.2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.seed_projector._net.2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.projectors.0._net.0.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.projectors.0._net.0.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.projectors.0._net.2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.projectors.0._net.2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.projectors.1._net.0.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.projectors.1._net.0.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.projectors.1._net.2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.projectors.1._net.2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.projectors.2._net.0.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.projectors.2._net.0.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.projectors.2._net.2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.projectors.2._net.2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.projectors.3._net.0.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.projectors.3._net.0.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.projectors.3._net.2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.projectors.3._net.2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.attractors.0._net.0.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.attractors.0._net.0.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.attractors.0._net.2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.attractors.0._net.2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.attractors.1._net.0.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.attractors.1._net.0.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.attractors.1._net.2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.attractors.1._net.2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.attractors.2._net.0.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.attractors.2._net.0.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.attractors.2._net.2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.attractors.2._net.2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.attractors.3._net.0.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.attractors.3._net.0.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.attractors.3._net.2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.attractors.3._net.2.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.conditional_log_binomial.mlp.0.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.conditional_log_binomial.mlp.0.bias
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.conditional_log_binomial.mlp.2.weight
+2024/03/15 09:55:34 - patchstitcher - INFO - training param: module.coarse_branch.conditional_log_binomial.mlp.2.bias
+2024/03/15 09:57:50 - patchstitcher - INFO - Epoch: [01/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.9729490280151367 - coarse_loss: 1.9729490280151367
+2024/03/15 09:59:39 - patchstitcher - INFO - Epoch: [01/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.6159499883651733 - coarse_loss: 1.6159499883651733
+2024/03/15 10:01:20 - patchstitcher - INFO - Epoch: [01/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.6653645038604736 - coarse_loss: 1.6653645038604736
+2024/03/15 10:03:08 - patchstitcher - INFO - Epoch: [01/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.3738189935684204 - coarse_loss: 1.3738189935684204
+2024/03/15 10:06:24 - patchstitcher - INFO - Epoch: [02/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.0679881572723389 - coarse_loss: 1.0679881572723389
+2024/03/15 10:08:12 - patchstitcher - INFO - Epoch: [02/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.0449714660644531 - coarse_loss: 1.0449714660644531
+2024/03/15 10:09:57 - patchstitcher - INFO - Epoch: [02/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.3200674057006836 - coarse_loss: 1.3200674057006836
+2024/03/15 10:11:44 - patchstitcher - INFO - Epoch: [02/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.2463884353637695 - coarse_loss: 1.2463884353637695
+2024/03/15 10:13:21 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+----------+----------+-----------+-----------+-----------+-----------+-----------+-----------+
+|     a1    |     a2    |    a3    | abs_rel  |    rmse   |   log_10  |  rmse_log |   silog   |   sq_rel  |    see    |
++-----------+-----------+----------+----------+-----------+-----------+-----------+-----------+-----------+-----------+
+| 0.9277873 | 0.9864464 | 0.994876 | 0.093889 | 1.7125608 | 0.0411139 | 0.1284599 | 10.310956 | 0.2504752 | 1.2484615 |
++-----------+-----------+----------+----------+-----------+-----------+-----------+-----------+-----------+-----------+
+2024/03/15 10:15:11 - patchstitcher - INFO - Epoch: [03/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.1642838716506958 - coarse_loss: 1.1642838716506958
+2024/03/15 10:16:56 - patchstitcher - INFO - Epoch: [03/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.1062591075897217 - coarse_loss: 1.1062591075897217
+2024/03/15 10:18:40 - patchstitcher - INFO - Epoch: [03/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.491640329360962 - coarse_loss: 1.491640329360962
+2024/03/15 10:20:26 - patchstitcher - INFO - Epoch: [03/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.0693362951278687 - coarse_loss: 1.0693362951278687
+2024/03/15 10:23:28 - patchstitcher - INFO - Epoch: [04/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.2830930948257446 - coarse_loss: 1.2830930948257446
+2024/03/15 10:25:13 - patchstitcher - INFO - Epoch: [04/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.8494630455970764 - coarse_loss: 0.8494630455970764
+2024/03/15 10:26:59 - patchstitcher - INFO - Epoch: [04/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.100481390953064 - coarse_loss: 1.100481390953064
+2024/03/15 10:28:45 - patchstitcher - INFO - Epoch: [04/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.6722239255905151 - coarse_loss: 0.6722239255905151
+2024/03/15 10:30:18 - patchstitcher - INFO - Evaluation Summary: 
++----------+-----------+-----------+-----------+-----------+----------+-----------+-----------+-----------+-----------+
+|    a1    |     a2    |     a3    |  abs_rel  |    rmse   |  log_10  |  rmse_log |   silog   |   sq_rel  |    see    |
++----------+-----------+-----------+-----------+-----------+----------+-----------+-----------+-----------+-----------+
+| 0.961338 | 0.9893523 | 0.9953463 | 0.0692743 | 1.5390607 | 0.030108 | 0.1050118 | 9.1967623 | 0.1975309 | 1.1110629 |
++----------+-----------+-----------+-----------+-----------+----------+-----------+-----------+-----------+-----------+
+2024/03/15 10:32:10 - patchstitcher - INFO - Epoch: [05/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.5996298789978027 - coarse_loss: 0.5996298789978027
+2024/03/15 10:33:58 - patchstitcher - INFO - Epoch: [05/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.5094271302223206 - coarse_loss: 0.5094271302223206
+2024/03/15 10:35:48 - patchstitcher - INFO - Epoch: [05/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.7459169626235962 - coarse_loss: 0.7459169626235962
+2024/03/15 10:37:33 - patchstitcher - INFO - Epoch: [05/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.7367539405822754 - coarse_loss: 0.7367539405822754
+2024/03/15 10:40:39 - patchstitcher - INFO - Epoch: [06/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.3089935779571533 - coarse_loss: 1.3089935779571533
+2024/03/15 10:42:24 - patchstitcher - INFO - Epoch: [06/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.9458222985267639 - coarse_loss: 0.9458222985267639
+2024/03/15 10:44:12 - patchstitcher - INFO - Epoch: [06/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.7383743524551392 - coarse_loss: 0.7383743524551392
+2024/03/15 10:45:59 - patchstitcher - INFO - Epoch: [06/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.6774943470954895 - coarse_loss: 0.6774943470954895
+2024/03/15 10:47:29 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+----------+----------+-----------+-----------+-----------+-----------+-----------+
+|     a1    |     a2    |     a3    | abs_rel  |   rmse   |   log_10  |  rmse_log |   silog   |   sq_rel  |    see    |
++-----------+-----------+-----------+----------+----------+-----------+-----------+-----------+-----------+-----------+
+| 0.9625513 | 0.9896059 | 0.9953454 | 0.076086 | 1.553624 | 0.0339274 | 0.1113379 | 8.9179546 | 0.1912439 | 1.0962123 |
++-----------+-----------+-----------+----------+----------+-----------+-----------+-----------+-----------+-----------+
+2024/03/15 10:49:20 - patchstitcher - INFO - Epoch: [07/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.7863880395889282 - coarse_loss: 0.7863880395889282
+2024/03/15 10:51:04 - patchstitcher - INFO - Epoch: [07/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.1585361957550049 - coarse_loss: 1.1585361957550049
+2024/03/15 10:52:54 - patchstitcher - INFO - Epoch: [07/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.1414254903793335 - coarse_loss: 1.1414254903793335
+2024/03/15 10:54:41 - patchstitcher - INFO - Epoch: [07/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.6607706546783447 - coarse_loss: 0.6607706546783447
+2024/03/15 10:57:47 - patchstitcher - INFO - Epoch: [08/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.8438395857810974 - coarse_loss: 0.8438395857810974
+2024/03/15 10:59:37 - patchstitcher - INFO - Epoch: [08/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.931841254234314 - coarse_loss: 0.931841254234314
+2024/03/15 11:01:23 - patchstitcher - INFO - Epoch: [08/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.2649768590927124 - coarse_loss: 1.2649768590927124
+2024/03/15 11:03:05 - patchstitcher - INFO - Epoch: [08/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.356317400932312 - coarse_loss: 1.356317400932312
+2024/03/15 11:04:39 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  |  rmse_log |   silog   |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+| 0.9688624 | 0.9900475 | 0.9955412 | 0.0621825 | 1.4741381 | 0.0269014 | 0.0983563 | 8.5882915 | 0.1738514 | 1.0249666 |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+2024/03/15 11:06:28 - patchstitcher - INFO - Epoch: [09/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.1434571743011475 - coarse_loss: 1.1434571743011475
+2024/03/15 11:08:19 - patchstitcher - INFO - Epoch: [09/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.7681660652160645 - coarse_loss: 1.7681660652160645
+2024/03/15 11:10:04 - patchstitcher - INFO - Epoch: [09/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.8547622561454773 - coarse_loss: 0.8547622561454773
+2024/03/15 11:11:49 - patchstitcher - INFO - Epoch: [09/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.869714617729187 - coarse_loss: 0.869714617729187
+2024/03/15 11:14:59 - patchstitcher - INFO - Epoch: [10/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.5332772731781006 - coarse_loss: 0.5332772731781006
+2024/03/15 11:16:44 - patchstitcher - INFO - Epoch: [10/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.8691495060920715 - coarse_loss: 0.8691495060920715
+2024/03/15 11:18:28 - patchstitcher - INFO - Epoch: [10/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.4371870756149292 - coarse_loss: 1.4371870756149292
+2024/03/15 11:20:14 - patchstitcher - INFO - Epoch: [10/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.9575653076171875 - coarse_loss: 0.9575653076171875
+2024/03/15 11:21:45 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  |  rmse_log |   silog   |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+| 0.9679335 | 0.9903103 | 0.9957452 | 0.0634565 | 1.4144222 | 0.0269387 | 0.0964634 | 8.5336222 | 0.1681394 | 1.0266862 |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+2024/03/15 11:23:36 - patchstitcher - INFO - Epoch: [11/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.8048105835914612 - coarse_loss: 0.8048105835914612
+2024/03/15 11:25:22 - patchstitcher - INFO - Epoch: [11/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.8616613149642944 - coarse_loss: 0.8616613149642944
+2024/03/15 11:27:12 - patchstitcher - INFO - Epoch: [11/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.221915364265442 - coarse_loss: 1.221915364265442
+2024/03/15 11:28:59 - patchstitcher - INFO - Epoch: [11/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.5273403525352478 - coarse_loss: 0.5273403525352478
+2024/03/15 11:31:59 - patchstitcher - INFO - Epoch: [12/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.6490796208381653 - coarse_loss: 0.6490796208381653
+2024/03/15 11:33:46 - patchstitcher - INFO - Epoch: [12/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.9228641986846924 - coarse_loss: 0.9228641986846924
+2024/03/15 11:35:30 - patchstitcher - INFO - Epoch: [12/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.8991217017173767 - coarse_loss: 0.8991217017173767
+2024/03/15 11:37:21 - patchstitcher - INFO - Epoch: [12/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.778996467590332 - coarse_loss: 0.778996467590332
+2024/03/15 11:38:51 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+-----------+----------+-----------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  | rmse_log |   silog   |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+-----------+-----------+----------+-----------+-----------+-----------+
+| 0.9700605 | 0.9907225 | 0.9956863 | 0.0593423 | 1.3817834 | 0.0258237 | 0.095056 | 8.4508466 | 0.1639893 | 1.0006335 |
++-----------+-----------+-----------+-----------+-----------+-----------+----------+-----------+-----------+-----------+
+2024/03/15 11:40:42 - patchstitcher - INFO - Epoch: [13/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.2246499061584473 - coarse_loss: 1.2246499061584473
+2024/03/15 11:42:33 - patchstitcher - INFO - Epoch: [13/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.055445671081543 - coarse_loss: 1.055445671081543
+2024/03/15 11:44:18 - patchstitcher - INFO - Epoch: [13/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.8403045535087585 - coarse_loss: 0.8403045535087585
+2024/03/15 11:46:03 - patchstitcher - INFO - Epoch: [13/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.7852007150650024 - coarse_loss: 0.7852007150650024
+2024/03/15 11:49:05 - patchstitcher - INFO - Epoch: [14/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.5313113331794739 - coarse_loss: 0.5313113331794739
+2024/03/15 11:50:53 - patchstitcher - INFO - Epoch: [14/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.803260326385498 - coarse_loss: 0.803260326385498
+2024/03/15 11:52:35 - patchstitcher - INFO - Epoch: [14/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.6353864669799805 - coarse_loss: 0.6353864669799805
+2024/03/15 11:54:22 - patchstitcher - INFO - Epoch: [14/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.603277862071991 - coarse_loss: 0.603277862071991
+2024/03/15 11:55:54 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+----------+-----------+-----------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |  log_10  |  rmse_log |   silog   |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+-----------+----------+-----------+-----------+-----------+-----------+
+| 0.9716077 | 0.9908243 | 0.9958379 | 0.0603097 | 1.3795547 | 0.025826 | 0.0942337 | 8.2481922 | 0.1615328 | 1.0314286 |
++-----------+-----------+-----------+-----------+-----------+----------+-----------+-----------+-----------+-----------+
+2024/03/15 11:57:46 - patchstitcher - INFO - Epoch: [15/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.68681800365448 - coarse_loss: 0.68681800365448
+2024/03/15 11:59:38 - patchstitcher - INFO - Epoch: [15/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.8562105894088745 - coarse_loss: 0.8562105894088745
+2024/03/15 12:01:24 - patchstitcher - INFO - Epoch: [15/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.0672423839569092 - coarse_loss: 1.0672423839569092
+2024/03/15 12:03:05 - patchstitcher - INFO - Epoch: [15/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.7026287317276001 - coarse_loss: 0.7026287317276001
+2024/03/15 12:06:08 - patchstitcher - INFO - Epoch: [16/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.9886091947555542 - coarse_loss: 0.9886091947555542
+2024/03/15 12:07:54 - patchstitcher - INFO - Epoch: [16/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.6522326469421387 - coarse_loss: 0.6522326469421387
+2024/03/15 12:09:39 - patchstitcher - INFO - Epoch: [16/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.9577221870422363 - coarse_loss: 0.9577221870422363
+2024/03/15 12:11:22 - patchstitcher - INFO - Epoch: [16/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.7307658195495605 - coarse_loss: 1.7307658195495605
+2024/03/15 12:12:51 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  |  rmse_log |   silog   |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+| 0.9747152 | 0.9908944 | 0.9959154 | 0.0511857 | 1.3574797 | 0.0221211 | 0.0867927 | 7.9538576 | 0.1511261 | 1.0003225 |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+2024/03/15 12:14:43 - patchstitcher - INFO - Epoch: [17/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.5646010637283325 - coarse_loss: 0.5646010637283325
+2024/03/15 12:16:28 - patchstitcher - INFO - Epoch: [17/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.8057535290718079 - coarse_loss: 0.8057535290718079
+2024/03/15 12:18:17 - patchstitcher - INFO - Epoch: [17/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.11107337474823 - coarse_loss: 1.11107337474823
+2024/03/15 12:20:01 - patchstitcher - INFO - Epoch: [17/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.187990427017212 - coarse_loss: 1.187990427017212
+2024/03/15 12:23:09 - patchstitcher - INFO - Epoch: [18/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.6382083892822266 - coarse_loss: 0.6382083892822266
+2024/03/15 12:24:49 - patchstitcher - INFO - Epoch: [18/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.5392951965332031 - coarse_loss: 0.5392951965332031
+2024/03/15 12:26:37 - patchstitcher - INFO - Epoch: [18/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.8188748359680176 - coarse_loss: 0.8188748359680176
+2024/03/15 12:28:18 - patchstitcher - INFO - Epoch: [18/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.0811210870742798 - coarse_loss: 1.0811210870742798
+2024/03/15 12:29:49 - patchstitcher - INFO - Evaluation Summary: 
++----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+|    a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  |  rmse_log |   silog   |   sq_rel  |    see    |
++----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+| 0.975323 | 0.9911468 | 0.9959684 | 0.0483459 | 1.3259571 | 0.0207656 | 0.0842995 | 7.8959624 | 0.1478599 | 0.9762505 |
++----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+2024/03/15 12:31:43 - patchstitcher - INFO - Epoch: [19/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.5005310773849487 - coarse_loss: 0.5005310773849487
+2024/03/15 12:33:28 - patchstitcher - INFO - Epoch: [19/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.5474035143852234 - coarse_loss: 0.5474035143852234
+2024/03/15 12:35:16 - patchstitcher - INFO - Epoch: [19/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.7799822092056274 - coarse_loss: 0.7799822092056274
+2024/03/15 12:37:02 - patchstitcher - INFO - Epoch: [19/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.5381927490234375 - coarse_loss: 0.5381927490234375
+2024/03/15 12:40:07 - patchstitcher - INFO - Epoch: [20/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.1203773021697998 - coarse_loss: 1.1203773021697998
+2024/03/15 12:41:51 - patchstitcher - INFO - Epoch: [20/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.5552318096160889 - coarse_loss: 0.5552318096160889
+2024/03/15 12:43:35 - patchstitcher - INFO - Epoch: [20/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.4946790933609009 - coarse_loss: 0.4946790933609009
+2024/03/15 12:45:21 - patchstitcher - INFO - Epoch: [20/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.829839825630188 - coarse_loss: 0.829839825630188
+2024/03/15 12:46:50 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+----------+-----------+-----------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |  log_10  |  rmse_log |   silog   |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+-----------+----------+-----------+-----------+-----------+-----------+
+| 0.9759003 | 0.9912674 | 0.9959566 | 0.0472804 | 1.3156906 | 0.020464 | 0.0841626 | 7.7711489 | 0.1448604 | 0.9643456 |
++-----------+-----------+-----------+-----------+-----------+----------+-----------+-----------+-----------+-----------+
+2024/03/15 12:48:43 - patchstitcher - INFO - Epoch: [21/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.8187640905380249 - coarse_loss: 0.8187640905380249
+2024/03/15 12:50:30 - patchstitcher - INFO - Epoch: [21/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.5510168671607971 - coarse_loss: 0.5510168671607971
+2024/03/15 12:52:22 - patchstitcher - INFO - Epoch: [21/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.5071703791618347 - coarse_loss: 0.5071703791618347
+2024/03/15 12:54:08 - patchstitcher - INFO - Epoch: [21/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.6241310834884644 - coarse_loss: 1.6241310834884644
+2024/03/15 12:57:18 - patchstitcher - INFO - Epoch: [22/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.9662288427352905 - coarse_loss: 0.9662288427352905
+2024/03/15 12:59:03 - patchstitcher - INFO - Epoch: [22/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.915822446346283 - coarse_loss: 0.915822446346283
+2024/03/15 13:00:45 - patchstitcher - INFO - Epoch: [22/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.48746258020401 - coarse_loss: 0.48746258020401
+2024/03/15 13:02:29 - patchstitcher - INFO - Epoch: [22/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.7346612811088562 - coarse_loss: 0.7346612811088562
+2024/03/15 13:04:01 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+----------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  |  rmse_log |  silog   |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+----------+-----------+-----------+
+| 0.9762025 | 0.9913185 | 0.9959977 | 0.0456843 | 1.3065255 | 0.0197035 | 0.0823783 | 7.684332 | 0.1431234 | 0.9606835 |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+----------+-----------+-----------+
+2024/03/15 13:05:51 - patchstitcher - INFO - Epoch: [23/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.5825149416923523 - coarse_loss: 0.5825149416923523
+2024/03/15 13:07:38 - patchstitcher - INFO - Epoch: [23/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.0635181665420532 - coarse_loss: 1.0635181665420532
+2024/03/15 13:09:24 - patchstitcher - INFO - Epoch: [23/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.632516622543335 - coarse_loss: 1.632516622543335
+2024/03/15 13:11:08 - patchstitcher - INFO - Epoch: [23/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.353378415107727 - coarse_loss: 1.353378415107727
+2024/03/15 13:14:18 - patchstitcher - INFO - Epoch: [24/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.8277870416641235 - coarse_loss: 0.8277870416641235
+2024/03/15 13:16:02 - patchstitcher - INFO - Epoch: [24/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.5105581283569336 - coarse_loss: 0.5105581283569336
+2024/03/15 13:17:45 - patchstitcher - INFO - Epoch: [24/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.43523621559143066 - coarse_loss: 0.43523621559143066
+2024/03/15 13:19:31 - patchstitcher - INFO - Epoch: [24/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.40485745668411255 - coarse_loss: 0.40485745668411255
+2024/03/15 13:21:02 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  |  rmse_log |   silog   |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+| 0.9762546 | 0.9913396 | 0.9959976 | 0.0452784 | 1.2974494 | 0.0194901 | 0.0821238 | 7.7005432 | 0.1431584 | 0.9635146 |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+2024/03/15 13:21:02 - patchstitcher - INFO - Saving ckp, but use the inner get_save_dict fuction to get model_dict
+2024/03/15 13:21:02 - patchstitcher - INFO - For saving space. Would you like to save base model several times? :>
+2024/03/15 13:21:03 - patchstitcher - INFO - save checkpoint_24.pth at ./work_dir/depthanything_vitb_u4k/coarse_pretrain

depthanything_vitb_u4k/coarse_pretrain/checkpoint_24.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4fa7eccecb3ba6b7f7e7aabcb8e1cc7be703da3d6eaff316bf22237a616b2afb
+size 1171453994

depthanything_vitb_u4k/coarse_pretrain/config.py

@@ -0,0 +1,310 @@
+collect_input_args = [
+    'image_lr',
+    'crops_image_hr',
+    'depth_gt',
+    'crop_depths',
+    'bboxs',
+    'image_hr',
+]
+convert_syncbn = True
+debug = False
+env_cfg = dict(
+    cudnn_benchmark=True,
+    dist_cfg=dict(backend='nccl'),
+    mp_cfg=dict(mp_start_method='forkserver'))
+find_unused_parameters = True
+general_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        dataset_name='', gt_dir=None, rgb_image_dir='', type='ImageDataset'),
+    num_workers=2)
+launcher = 'pytorch'
+log_name = 'coarse_pretrain'
+max_depth = 80
+min_depth = 0.001
+model = dict(
+    coarse_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        depth_anything=True,
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            392,
+            518,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='vitb',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource='local::./work_dir/DepthAnything_vitb.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='DA-ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    fine_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        depth_anything=True,
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            392,
+            518,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='vitb',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource='local::./work_dir/DepthAnything_vitb.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='DA-ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    max_depth=80,
+    min_depth=0.001,
+    sigloss=dict(type='SILogLoss'),
+    target='coarse',
+    type='BaselinePretrain')
+optim_wrapper = dict(
+    clip_grad=dict(max_norm=0.1, norm_type=2, type='norm'),
+    optimizer=dict(lr=4e-06, type='AdamW', weight_decay=0.01),
+    paramwise_cfg=dict(bypass_duplicate=True, custom_keys=dict()))
+param_scheduler = dict(
+    base_momentum=0.85,
+    cycle_momentum=True,
+    div_factor=1,
+    final_div_factor=10000,
+    max_momentum=0.95,
+    pct_start=0.5,
+    three_phase=False)
+project = 'patchfusion'
+resume = False
+tags = [
+    'coarse',
+    'da',
+    'vitb',
+]
+test_in_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+test_out_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test_out.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+train_cfg = dict(
+    eval_start=0,
+    log_interval=100,
+    max_epochs=24,
+    save_checkpoint_interval=24,
+    train_log_img_interval=500,
+    val_interval=2,
+    val_log_img_interval=50,
+    val_type='epoch_base')
+train_dataloader = dict(
+    batch_size=4,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='train',
+        resize_mode='depth-anything',
+        split='./data/u4k/splits/train.txt',
+        transform_cfg=dict(
+            degree=1.0,
+            network_process_size=[
+                392,
+                518,
+            ],
+            random_crop=True,
+            random_crop_size=(
+                540,
+                960,
+            )),
+        type='UnrealStereo4kDataset'),
+    num_workers=4)
+val_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        resize_mode='depth-anything',
+        split='./data/u4k/splits/val.txt',
+        transform_cfg=dict(
+            degree=1.0,
+            network_process_size=[
+                392,
+                518,
+            ],
+            random_crop_size=(
+                540,
+                960,
+            )),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+work_dir = './work_dir/depthanything_vitb_u4k/coarse_pretrain'
+zoe_depth_config = dict(
+    attractor_alpha=1000,
+    attractor_gamma=2,
+    attractor_kind='mean',
+    attractor_type='inv',
+    aug=True,
+    bin_centers_type='softplus',
+    bin_embedding_dim=128,
+    clip_grad=0.1,
+    dataset='nyu',
+    depth_anything=True,
+    distributed=True,
+    do_resize=False,
+    force_keep_ar=True,
+    freeze_midas_bn=True,
+    gpu='NULL',
+    img_size=[
+        392,
+        518,
+    ],
+    inverse_midas=False,
+    log_images_every=0.1,
+    max_depth=80,
+    max_temp=50.0,
+    max_translation=100,
+    memory_efficient=True,
+    midas_model_type='vitb',
+    min_depth=0.001,
+    min_temp=0.0212,
+    model='zoedepth',
+    n_attractors=[
+        16,
+        8,
+        4,
+        1,
+    ],
+    n_bins=64,
+    name='ZoeDepth',
+    notes='',
+    output_distribution='logbinomial',
+    prefetch=False,
+    pretrained_resource='local::./work_dir/DepthAnything_vitb.pt',
+    print_losses=False,
+    project='ZoeDepth',
+    random_crop=False,
+    random_translate=False,
+    root='.',
+    save_dir='',
+    shared_dict='NULL',
+    tags='',
+    train_midas=True,
+    translate_prob=0.2,
+    type='DA-ZoeDepth',
+    uid='NULL',
+    use_amp=False,
+    use_pretrained_midas=True,
+    use_shared_dict=False,
+    validate_every=0.25,
+    version_name='v1',
+    workers=16)

depthanything_vitb_u4k/fine_pretrain/20240315_153036.log

@@ -0,0 +1,1028 @@
+2024/03/15 15:30:44 - patchstitcher - INFO - 
+------------------------------------------------------------
+System environment:
+    sys.platform: linux
+    Python: 3.8.18 | packaged by conda-forge | (default, Oct 10 2023, 15:44:36) [GCC 12.3.0]
+    CUDA available: True
+    numpy_random_seed: 621
+    GPU 0,1,2,3: NVIDIA A100-SXM4-80GB
+    CUDA_HOME: /sw/rl9g/cuda/11.8/rl9_binary
+    NVCC: Cuda compilation tools, release 11.8, V11.8.89
+    GCC: gcc (GCC) 11.3.1 20220421 (Red Hat 11.3.1-2)
+    PyTorch: 2.1.2
+    PyTorch compiling details: PyTorch built with:
+  - GCC 9.3
+  - C++ Version: 201703
+  - Intel(R) oneAPI Math Kernel Library Version 2022.1-Product Build 20220311 for Intel(R) 64 architecture applications
+  - Intel(R) MKL-DNN v3.1.1 (Git Hash 64f6bcbcbab628e96f33a62c3e975f8535a7bde4)
+  - OpenMP 201511 (a.k.a. OpenMP 4.5)
+  - LAPACK is enabled (usually provided by MKL)
+  - NNPACK is enabled
+  - CPU capability usage: AVX2
+  - CUDA Runtime 11.8
+  - NVCC architecture flags: -gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_61,code=sm_61;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_80,code=sm_80;-gencode;arch=compute_86,code=sm_86;-gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_90,code=sm_90;-gencode;arch=compute_37,code=compute_37
+  - CuDNN 8.7
+  - Magma 2.6.1
+  - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=11.8, CUDNN_VERSION=8.7.0, CXX_COMPILER=/opt/rh/devtoolset-9/root/usr/bin/c++, CXX_FLAGS= -D_GLIBCXX_USE_CXX11_ABI=0 -fabi-version=11 -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOROCTRACER -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -O2 -fPIC -Wall -Wextra -Werror=return-type -Werror=non-virtual-dtor -Werror=bool-operation -Wnarrowing -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wno-unused-parameter -Wno-unused-function -Wno-unused-result -Wno-strict-overflow -Wno-strict-aliasing -Wno-stringop-overflow -Wno-psabi -Wno-error=pedantic -Wno-error=old-style-cast -Wno-invalid-partial-specialization -Wno-unused-private-field -Wno-aligned-allocation-unavailable -Wno-missing-braces -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Werror=cast-function-type -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_DISABLE_GPU_ASSERTS=ON, TORCH_VERSION=2.1.2, USE_CUDA=ON, USE_CUDNN=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=ON, USE_NNPACK=ON, USE_OPENMP=ON, USE_ROCM=OFF, 
+
+    TorchVision: 0.16.2
+    OpenCV: 4.8.1
+    MMEngine: 0.10.2
+
+Runtime environment:
+    cudnn_benchmark: True
+    mp_cfg: {'mp_start_method': 'forkserver'}
+    dist_cfg: {'backend': 'nccl'}
+    seed: 621
+    Distributed launcher: pytorch
+    Distributed training: True
+    GPU number: 4
+------------------------------------------------------------
+
+2024/03/15 15:30:44 - patchstitcher - INFO - Config:
+collect_input_args = [
+    'image_lr',
+    'crops_image_hr',
+    'depth_gt',
+    'crop_depths',
+    'bboxs',
+    'image_hr',
+]
+convert_syncbn = True
+debug = False
+env_cfg = dict(
+    cudnn_benchmark=True,
+    dist_cfg=dict(backend='nccl'),
+    mp_cfg=dict(mp_start_method='forkserver'))
+find_unused_parameters = True
+general_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        dataset_name='', gt_dir=None, rgb_image_dir='', type='ImageDataset'),
+    num_workers=2)
+launcher = 'pytorch'
+log_name = 'fine_pretrain'
+max_depth = 80
+min_depth = 0.001
+model = dict(
+    coarse_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        depth_anything=True,
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            392,
+            518,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='vitb',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource='local::./work_dir/DepthAnything_vitb.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='DA-ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    fine_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        depth_anything=True,
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            392,
+            518,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='vitb',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource='local::./work_dir/DepthAnything_vitb.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='DA-ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    max_depth=80,
+    min_depth=0.001,
+    patch_process_shape=(
+        392,
+        518,
+    ),
+    sigloss=dict(type='SILogLoss'),
+    target='fine',
+    type='BaselinePretrain')
+optim_wrapper = dict(
+    clip_grad=dict(max_norm=0.1, norm_type=2, type='norm'),
+    optimizer=dict(lr=4e-06, type='AdamW', weight_decay=0.01),
+    paramwise_cfg=dict(bypass_duplicate=True, custom_keys=dict()))
+param_scheduler = dict(
+    base_momentum=0.85,
+    cycle_momentum=True,
+    div_factor=1,
+    final_div_factor=10000,
+    max_momentum=0.95,
+    pct_start=0.5,
+    three_phase=False)
+project = 'patchfusion'
+tags = [
+    'fine',
+    'da',
+    'vitb',
+]
+test_in_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+test_out_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test_out.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+train_cfg = dict(
+    eval_start=0,
+    log_interval=100,
+    max_epochs=24,
+    save_checkpoint_interval=24,
+    train_log_img_interval=500,
+    val_interval=2,
+    val_log_img_interval=50,
+    val_type='epoch_base')
+train_dataloader = dict(
+    batch_size=4,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='train',
+        resize_mode='depth-anything',
+        split='./data/u4k/splits/train.txt',
+        transform_cfg=dict(
+            degree=1.0, network_process_size=[
+                392,
+                518,
+            ], random_crop=True),
+        type='UnrealStereo4kDataset'),
+    num_workers=4)
+val_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        resize_mode='depth-anything',
+        split='./data/u4k/splits/val.txt',
+        transform_cfg=dict(degree=1.0, network_process_size=[
+            392,
+            518,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+work_dir = './work_dir/depthanything_vitb_u4k/fine_pretrain'
+zoe_depth_config = dict(
+    attractor_alpha=1000,
+    attractor_gamma=2,
+    attractor_kind='mean',
+    attractor_type='inv',
+    aug=True,
+    bin_centers_type='softplus',
+    bin_embedding_dim=128,
+    clip_grad=0.1,
+    dataset='nyu',
+    depth_anything=True,
+    distributed=True,
+    do_resize=False,
+    force_keep_ar=True,
+    freeze_midas_bn=True,
+    gpu='NULL',
+    img_size=[
+        392,
+        518,
+    ],
+    inverse_midas=False,
+    log_images_every=0.1,
+    max_depth=80,
+    max_temp=50.0,
+    max_translation=100,
+    memory_efficient=True,
+    midas_model_type='vitb',
+    min_depth=0.001,
+    min_temp=0.0212,
+    model='zoedepth',
+    n_attractors=[
+        16,
+        8,
+        4,
+        1,
+    ],
+    n_bins=64,
+    name='ZoeDepth',
+    notes='',
+    output_distribution='logbinomial',
+    prefetch=False,
+    pretrained_resource='local::./work_dir/DepthAnything_vitb.pt',
+    print_losses=False,
+    project='ZoeDepth',
+    random_crop=False,
+    random_translate=False,
+    root='.',
+    save_dir='',
+    shared_dict='NULL',
+    tags='',
+    train_midas=True,
+    translate_prob=0.2,
+    type='DA-ZoeDepth',
+    uid='NULL',
+    use_amp=False,
+    use_pretrained_midas=True,
+    use_shared_dict=False,
+    validate_every=0.25,
+    version_name='v1',
+    workers=16)
+
+2024/03/15 15:30:45 - patchstitcher - INFO - Loading deepnet from local::./work_dir/DepthAnything_vitb.pt
+2024/03/15 15:30:45 - patchstitcher - INFO - Current zoedepth.core.prep.resizer is <class 'torch.nn.modules.linear.Identity'>
+2024/03/15 15:30:45 - patchstitcher - INFO - DistributedDataParallel(
+  (module): BaselinePretrain(
+    (fine_branch): ZoeDepth(
+      (core): DepthAnythingCore(
+        (core): DPT_DINOv2(
+          (pretrained): DinoVisionTransformer(
+            (patch_embed): PatchEmbed(
+              (proj): Conv2d(3, 768, kernel_size=(14, 14), stride=(14, 14))
+              (norm): Identity()
+            )
+            (blocks): ModuleList(
+              (0-11): 12 x NestedTensorBlock(
+                (norm1): LayerNorm((768,), eps=1e-06, elementwise_affine=True)
+                (attn): MemEffAttention(
+                  (qkv): Linear(in_features=768, out_features=2304, bias=True)
+                  (attn_drop): Dropout(p=0.0, inplace=False)
+                  (proj): Linear(in_features=768, out_features=768, bias=True)
+                  (proj_drop): Dropout(p=0.0, inplace=False)
+                )
+                (ls1): LayerScale()
+                (drop_path1): Identity()
+                (norm2): LayerNorm((768,), eps=1e-06, elementwise_affine=True)
+                (mlp): Mlp(
+                  (fc1): Linear(in_features=768, out_features=3072, bias=True)
+                  (act): GELU(approximate='none')
+                  (fc2): Linear(in_features=3072, out_features=768, bias=True)
+                  (drop): Dropout(p=0.0, inplace=False)
+                )
+                (ls2): LayerScale()
+                (drop_path2): Identity()
+              )
+            )
+            (norm): LayerNorm((768,), eps=1e-06, elementwise_affine=True)
+            (head): Identity()
+          )
+          (depth_head): DPTHead(
+            (projects): ModuleList(
+              (0): Conv2d(768, 96, kernel_size=(1, 1), stride=(1, 1))
+              (1): Conv2d(768, 192, kernel_size=(1, 1), stride=(1, 1))
+              (2): Conv2d(768, 384, kernel_size=(1, 1), stride=(1, 1))
+              (3): Conv2d(768, 768, kernel_size=(1, 1), stride=(1, 1))
+            )
+            (resize_layers): ModuleList(
+              (0): ConvTranspose2d(96, 96, kernel_size=(4, 4), stride=(4, 4))
+              (1): ConvTranspose2d(192, 192, kernel_size=(2, 2), stride=(2, 2))
+              (2): Identity()
+              (3): Conv2d(768, 768, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1))
+            )
+            (scratch): Module(
+              (layer1_rn): Conv2d(96, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
+              (layer2_rn): Conv2d(192, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
+              (layer3_rn): Conv2d(384, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
+              (layer4_rn): Conv2d(768, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
+              (refinenet1): FeatureFusionBlock(
+                (out_conv): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+                (resConfUnit1): ResidualConvUnit(
+                  (conv1): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (resConfUnit2): ResidualConvUnit(
+                  (conv1): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (skip_add): FloatFunctional(
+                  (activation_post_process): Identity()
+                )
+              )
+              (refinenet2): FeatureFusionBlock(
+                (out_conv): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+                (resConfUnit1): ResidualConvUnit(
+                  (conv1): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (resConfUnit2): ResidualConvUnit(
+                  (conv1): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (skip_add): FloatFunctional(
+                  (activation_post_process): Identity()
+                )
+              )
+              (refinenet3): FeatureFusionBlock(
+                (out_conv): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+                (resConfUnit1): ResidualConvUnit(
+                  (conv1): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (resConfUnit2): ResidualConvUnit(
+                  (conv1): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (skip_add): FloatFunctional(
+                  (activation_post_process): Identity()
+                )
+              )
+              (refinenet4): FeatureFusionBlock(
+                (out_conv): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+                (resConfUnit1): ResidualConvUnit(
+                  (conv1): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (resConfUnit2): ResidualConvUnit(
+                  (conv1): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (skip_add): FloatFunctional(
+                  (activation_post_process): Identity()
+                )
+              )
+              (output_conv1): Conv2d(128, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+              (output_conv2): Sequential(
+                (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                (1): ReLU(inplace=True)
+                (2): Conv2d(32, 1, kernel_size=(1, 1), stride=(1, 1))
+                (3): ReLU(inplace=True)
+                (4): Identity()
+              )
+            )
+          )
+        )
+      )
+      (conv2): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+      (seed_bin_regressor): SeedBinRegressorUnnormed(
+        (_net): Sequential(
+          (0): Conv2d(128, 256, kernel_size=(1, 1), stride=(1, 1))
+          (1): ReLU(inplace=True)
+          (2): Conv2d(256, 64, kernel_size=(1, 1), stride=(1, 1))
+          (3): Softplus(beta=1, threshold=20)
+        )
+      )
+      (seed_projector): Projector(
+        (_net): Sequential(
+          (0): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+          (1): ReLU(inplace=True)
+          (2): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+        )
+      )
+      (projectors): ModuleList(
+        (0-3): 4 x Projector(
+          (_net): Sequential(
+            (0): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+            (1): ReLU(inplace=True)
+            (2): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+          )
+        )
+      )
+      (attractors): ModuleList(
+        (0): AttractorLayerUnnormed(
+          (_net): Sequential(
+            (0): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+            (1): ReLU(inplace=True)
+            (2): Conv2d(128, 16, kernel_size=(1, 1), stride=(1, 1))
+            (3): Softplus(beta=1, threshold=20)
+          )
+        )
+        (1): AttractorLayerUnnormed(
+          (_net): Sequential(
+            (0): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+            (1): ReLU(inplace=True)
+            (2): Conv2d(128, 8, kernel_size=(1, 1), stride=(1, 1))
+            (3): Softplus(beta=1, threshold=20)
+          )
+        )
+        (2): AttractorLayerUnnormed(
+          (_net): Sequential(
+            (0): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+            (1): ReLU(inplace=True)
+            (2): Conv2d(128, 4, kernel_size=(1, 1), stride=(1, 1))
+            (3): Softplus(beta=1, threshold=20)
+          )
+        )
+        (3): AttractorLayerUnnormed(
+          (_net): Sequential(
+            (0): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+            (1): ReLU(inplace=True)
+            (2): Conv2d(128, 1, kernel_size=(1, 1), stride=(1, 1))
+            (3): Softplus(beta=1, threshold=20)
+          )
+        )
+      )
+      (conditional_log_binomial): ConditionalLogBinomial(
+        (log_binomial_transform): LogBinomial()
+        (mlp): Sequential(
+          (0): Conv2d(161, 80, kernel_size=(1, 1), stride=(1, 1))
+          (1): GELU(approximate='none')
+          (2): Conv2d(80, 4, kernel_size=(1, 1), stride=(1, 1))
+          (3): Softplus(beta=1, threshold=20)
+        )
+      )
+    )
+    (sigloss): SILogLoss()
+  )
+)
+2024/03/15 15:30:51 - patchstitcher - INFO - successfully init trainer
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.cls_token
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.pos_embed
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.mask_token
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.patch_embed.proj.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.patch_embed.proj.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.0.norm1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.0.norm1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.0.attn.qkv.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.0.attn.qkv.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.0.attn.proj.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.0.attn.proj.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.0.ls1.gamma
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.0.norm2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.0.norm2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.0.mlp.fc1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.0.mlp.fc1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.0.mlp.fc2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.0.mlp.fc2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.0.ls2.gamma
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.1.norm1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.1.norm1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.1.attn.qkv.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.1.attn.qkv.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.1.attn.proj.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.1.attn.proj.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.1.ls1.gamma
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.1.norm2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.1.norm2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.1.mlp.fc1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.1.mlp.fc1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.1.mlp.fc2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.1.mlp.fc2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.1.ls2.gamma
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.2.norm1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.2.norm1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.2.attn.qkv.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.2.attn.qkv.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.2.attn.proj.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.2.attn.proj.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.2.ls1.gamma
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.2.norm2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.2.norm2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.2.mlp.fc1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.2.mlp.fc1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.2.mlp.fc2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.2.mlp.fc2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.2.ls2.gamma
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.3.norm1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.3.norm1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.3.attn.qkv.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.3.attn.qkv.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.3.attn.proj.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.3.attn.proj.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.3.ls1.gamma
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.3.norm2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.3.norm2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.3.mlp.fc1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.3.mlp.fc1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.3.mlp.fc2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.3.mlp.fc2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.3.ls2.gamma
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.4.norm1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.4.norm1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.4.attn.qkv.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.4.attn.qkv.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.4.attn.proj.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.4.attn.proj.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.4.ls1.gamma
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.4.norm2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.4.norm2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.4.mlp.fc1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.4.mlp.fc1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.4.mlp.fc2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.4.mlp.fc2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.4.ls2.gamma
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.5.norm1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.5.norm1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.5.attn.qkv.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.5.attn.qkv.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.5.attn.proj.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.5.attn.proj.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.5.ls1.gamma
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.5.norm2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.5.norm2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.5.mlp.fc1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.5.mlp.fc1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.5.mlp.fc2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.5.mlp.fc2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.5.ls2.gamma
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.6.norm1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.6.norm1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.6.attn.qkv.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.6.attn.qkv.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.6.attn.proj.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.6.attn.proj.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.6.ls1.gamma
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.6.norm2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.6.norm2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.6.mlp.fc1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.6.mlp.fc1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.6.mlp.fc2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.6.mlp.fc2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.6.ls2.gamma
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.7.norm1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.7.norm1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.7.attn.qkv.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.7.attn.qkv.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.7.attn.proj.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.7.attn.proj.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.7.ls1.gamma
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.7.norm2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.7.norm2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.7.mlp.fc1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.7.mlp.fc1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.7.mlp.fc2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.7.mlp.fc2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.7.ls2.gamma
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.8.norm1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.8.norm1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.8.attn.qkv.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.8.attn.qkv.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.8.attn.proj.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.8.attn.proj.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.8.ls1.gamma
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.8.norm2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.8.norm2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.8.mlp.fc1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.8.mlp.fc1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.8.mlp.fc2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.8.mlp.fc2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.8.ls2.gamma
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.9.norm1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.9.norm1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.9.attn.qkv.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.9.attn.qkv.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.9.attn.proj.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.9.attn.proj.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.9.ls1.gamma
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.9.norm2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.9.norm2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.9.mlp.fc1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.9.mlp.fc1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.9.mlp.fc2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.9.mlp.fc2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.9.ls2.gamma
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.10.norm1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.10.norm1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.10.attn.qkv.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.10.attn.qkv.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.10.attn.proj.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.10.attn.proj.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.10.ls1.gamma
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.10.norm2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.10.norm2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.10.mlp.fc1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.10.mlp.fc1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.10.mlp.fc2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.10.mlp.fc2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.10.ls2.gamma
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.11.norm1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.11.norm1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.11.attn.qkv.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.11.attn.qkv.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.11.attn.proj.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.11.attn.proj.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.11.ls1.gamma
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.11.norm2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.11.norm2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.11.mlp.fc1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.11.mlp.fc1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.11.mlp.fc2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.11.mlp.fc2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.11.ls2.gamma
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.norm.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.norm.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.projects.0.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.projects.0.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.projects.1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.projects.1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.projects.2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.projects.2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.projects.3.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.projects.3.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.resize_layers.0.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.resize_layers.0.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.resize_layers.1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.resize_layers.1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.resize_layers.3.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.resize_layers.3.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.layer1_rn.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.layer2_rn.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.layer3_rn.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.layer4_rn.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet1.out_conv.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet1.out_conv.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet1.resConfUnit1.conv1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet1.resConfUnit1.conv1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet1.resConfUnit1.conv2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet1.resConfUnit1.conv2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet1.resConfUnit2.conv1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet1.resConfUnit2.conv1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet1.resConfUnit2.conv2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet1.resConfUnit2.conv2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet2.out_conv.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet2.out_conv.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet2.resConfUnit1.conv1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet2.resConfUnit1.conv1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet2.resConfUnit1.conv2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet2.resConfUnit1.conv2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet2.resConfUnit2.conv1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet2.resConfUnit2.conv1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet2.resConfUnit2.conv2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet2.resConfUnit2.conv2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet3.out_conv.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet3.out_conv.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet3.resConfUnit1.conv1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet3.resConfUnit1.conv1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet3.resConfUnit1.conv2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet3.resConfUnit1.conv2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet3.resConfUnit2.conv1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet3.resConfUnit2.conv1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet3.resConfUnit2.conv2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet3.resConfUnit2.conv2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet4.out_conv.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet4.out_conv.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet4.resConfUnit1.conv1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet4.resConfUnit1.conv1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet4.resConfUnit1.conv2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet4.resConfUnit1.conv2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet4.resConfUnit2.conv1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet4.resConfUnit2.conv1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet4.resConfUnit2.conv2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet4.resConfUnit2.conv2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.output_conv1.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.output_conv1.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.output_conv2.0.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.output_conv2.0.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.output_conv2.2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.output_conv2.2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.conv2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.conv2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.seed_bin_regressor._net.0.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.seed_bin_regressor._net.0.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.seed_bin_regressor._net.2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.seed_bin_regressor._net.2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.seed_projector._net.0.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.seed_projector._net.0.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.seed_projector._net.2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.seed_projector._net.2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.projectors.0._net.0.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.projectors.0._net.0.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.projectors.0._net.2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.projectors.0._net.2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.projectors.1._net.0.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.projectors.1._net.0.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.projectors.1._net.2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.projectors.1._net.2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.projectors.2._net.0.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.projectors.2._net.0.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.projectors.2._net.2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.projectors.2._net.2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.projectors.3._net.0.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.projectors.3._net.0.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.projectors.3._net.2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.projectors.3._net.2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.attractors.0._net.0.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.attractors.0._net.0.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.attractors.0._net.2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.attractors.0._net.2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.attractors.1._net.0.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.attractors.1._net.0.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.attractors.1._net.2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.attractors.1._net.2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.attractors.2._net.0.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.attractors.2._net.0.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.attractors.2._net.2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.attractors.2._net.2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.attractors.3._net.0.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.attractors.3._net.0.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.attractors.3._net.2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.attractors.3._net.2.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.conditional_log_binomial.mlp.0.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.conditional_log_binomial.mlp.0.bias
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.conditional_log_binomial.mlp.2.weight
+2024/03/15 15:30:51 - patchstitcher - INFO - training param: module.fine_branch.conditional_log_binomial.mlp.2.bias
+2024/03/15 15:33:25 - patchstitcher - INFO - Epoch: [01/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 2.288588523864746 - fine_loss: 2.288588523864746
+2024/03/15 15:35:13 - patchstitcher - INFO - Epoch: [01/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.749260425567627 - fine_loss: 1.749260425567627
+2024/03/15 15:36:58 - patchstitcher - INFO - Epoch: [01/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 2.603142499923706 - fine_loss: 2.603142499923706
+2024/03/15 15:38:59 - patchstitcher - INFO - Epoch: [01/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 3.0235860347747803 - fine_loss: 3.0235860347747803
+2024/03/15 15:42:38 - patchstitcher - INFO - Epoch: [02/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 2.2628891468048096 - fine_loss: 2.2628891468048096
+2024/03/15 15:44:44 - patchstitcher - INFO - Epoch: [02/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 2.2125635147094727 - fine_loss: 2.2125635147094727
+2024/03/15 15:46:44 - patchstitcher - INFO - Epoch: [02/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.884977102279663 - fine_loss: 1.884977102279663
+2024/03/15 15:48:46 - patchstitcher - INFO - Epoch: [02/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 3.667808771133423 - fine_loss: 3.667808771133423
+2024/03/15 15:50:43 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+----------+-----------+-----------+------------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |   rmse   |   log_10  |  rmse_log |   silog    |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+----------+-----------+-----------+------------+-----------+-----------+
+| 0.7653929 | 0.9569647 | 0.9891034 | 0.1631364 | 2.063872 | 0.0675193 | 0.2015772 | 17.5721867 | 0.3284417 | 1.5396647 |
++-----------+-----------+-----------+-----------+----------+-----------+-----------+------------+-----------+-----------+
+2024/03/15 15:52:52 - patchstitcher - INFO - Epoch: [03/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.9002115726470947 - fine_loss: 1.9002115726470947
+2024/03/15 15:54:51 - patchstitcher - INFO - Epoch: [03/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.533200979232788 - fine_loss: 1.533200979232788
+2024/03/15 15:56:53 - patchstitcher - INFO - Epoch: [03/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.3708069324493408 - fine_loss: 1.3708069324493408
+2024/03/15 15:58:56 - patchstitcher - INFO - Epoch: [03/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.3536834716796875 - fine_loss: 1.3536834716796875
+2024/03/15 16:02:35 - patchstitcher - INFO - Epoch: [04/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.4067535400390625 - fine_loss: 1.4067535400390625
+2024/03/15 16:04:38 - patchstitcher - INFO - Epoch: [04/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.571197509765625 - fine_loss: 1.571197509765625
+2024/03/15 16:06:40 - patchstitcher - INFO - Epoch: [04/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 2.9749035835266113 - fine_loss: 2.9749035835266113
+2024/03/15 16:08:48 - patchstitcher - INFO - Epoch: [04/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.893333911895752 - fine_loss: 0.893333911895752
+2024/03/15 16:10:40 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+----------+-----------+-----------+-----------+----------+------------+-----------+-----------+
+|     a1    |     a2    |    a3    |  abs_rel  |    rmse   |   log_10  | rmse_log |   silog    |   sq_rel  |    see    |
++-----------+-----------+----------+-----------+-----------+-----------+----------+------------+-----------+-----------+
+| 0.8439181 | 0.9733375 | 0.992747 | 0.1316369 | 1.8230734 | 0.0558847 | 0.171333 | 15.4284363 | 0.2575101 | 1.3799866 |
++-----------+-----------+----------+-----------+-----------+-----------+----------+------------+-----------+-----------+
+2024/03/15 16:12:51 - patchstitcher - INFO - Epoch: [05/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.5204694271087646 - fine_loss: 1.5204694271087646
+2024/03/15 16:14:53 - patchstitcher - INFO - Epoch: [05/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.0538222789764404 - fine_loss: 1.0538222789764404
+2024/03/15 16:17:00 - patchstitcher - INFO - Epoch: [05/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.246050477027893 - fine_loss: 1.246050477027893
+2024/03/15 16:19:04 - patchstitcher - INFO - Epoch: [05/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.4139764308929443 - fine_loss: 1.4139764308929443
+2024/03/15 16:22:40 - patchstitcher - INFO - Epoch: [06/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.5990095138549805 - fine_loss: 1.5990095138549805
+2024/03/15 16:24:45 - patchstitcher - INFO - Epoch: [06/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.4719877243041992 - fine_loss: 1.4719877243041992
+2024/03/15 16:26:49 - patchstitcher - INFO - Epoch: [06/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.998321533203125 - fine_loss: 0.998321533203125
+2024/03/15 16:28:52 - patchstitcher - INFO - Epoch: [06/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.2637615203857422 - fine_loss: 1.2637615203857422
+2024/03/15 16:30:46 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  |  rmse_log |   silog    |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+| 0.8831826 | 0.9846013 | 0.9953048 | 0.1145366 | 1.6448599 | 0.0488564 | 0.1510406 | 14.0402038 | 0.2199031 | 1.3085128 |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+2024/03/15 16:32:53 - patchstitcher - INFO - Epoch: [07/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.65132737159729 - fine_loss: 1.65132737159729
+2024/03/15 16:34:56 - patchstitcher - INFO - Epoch: [07/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.4322144985198975 - fine_loss: 1.4322144985198975
+2024/03/15 16:37:04 - patchstitcher - INFO - Epoch: [07/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.034339427947998 - fine_loss: 1.034339427947998
+2024/03/15 16:39:08 - patchstitcher - INFO - Epoch: [07/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.0732086896896362 - fine_loss: 1.0732086896896362
+2024/03/15 16:42:43 - patchstitcher - INFO - Epoch: [08/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.3489086627960205 - fine_loss: 1.3489086627960205
+2024/03/15 16:44:47 - patchstitcher - INFO - Epoch: [08/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.4356486797332764 - fine_loss: 1.4356486797332764
+2024/03/15 16:46:50 - patchstitcher - INFO - Epoch: [08/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.6865524649620056 - fine_loss: 0.6865524649620056
+2024/03/15 16:48:50 - patchstitcher - INFO - Epoch: [08/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.4590085744857788 - fine_loss: 1.4590085744857788
+2024/03/15 16:50:41 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+-----------+----------+------------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  | rmse_log |   silog    |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+-----------+-----------+----------+------------+-----------+-----------+
+| 0.8921932 | 0.9874671 | 0.9972081 | 0.1083586 | 1.6257898 | 0.0457595 | 0.142043 | 12.7745355 | 0.2076856 | 1.2743567 |
++-----------+-----------+-----------+-----------+-----------+-----------+----------+------------+-----------+-----------+
+2024/03/15 16:52:44 - patchstitcher - INFO - Epoch: [09/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.008254885673523 - fine_loss: 1.008254885673523
+2024/03/15 16:54:54 - patchstitcher - INFO - Epoch: [09/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.8210620880126953 - fine_loss: 0.8210620880126953
+2024/03/15 16:56:55 - patchstitcher - INFO - Epoch: [09/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.8681334257125854 - fine_loss: 1.8681334257125854
+2024/03/15 16:58:59 - patchstitcher - INFO - Epoch: [09/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.9568914771080017 - fine_loss: 0.9568914771080017
+2024/03/15 17:02:34 - patchstitcher - INFO - Epoch: [10/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.5452194213867188 - fine_loss: 1.5452194213867188
+2024/03/15 17:04:40 - patchstitcher - INFO - Epoch: [10/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.9237810373306274 - fine_loss: 0.9237810373306274
+2024/03/15 17:06:43 - patchstitcher - INFO - Epoch: [10/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.4192367792129517 - fine_loss: 1.4192367792129517
+2024/03/15 17:08:47 - patchstitcher - INFO - Epoch: [10/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.1616711616516113 - fine_loss: 1.1616711616516113
+2024/03/15 17:10:40 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+----------+-----------+----------+-----------+-----------+------------+-----------+-----------+
+|     a1    |     a2    |    a3    |  abs_rel  |   rmse   |   log_10  |  rmse_log |   silog    |   sq_rel  |    see    |
++-----------+-----------+----------+-----------+----------+-----------+-----------+------------+-----------+-----------+
+| 0.9095374 | 0.9878494 | 0.996491 | 0.1000458 | 1.529536 | 0.0445519 | 0.1377915 | 12.2980782 | 0.1741764 | 1.1720957 |
++-----------+-----------+----------+-----------+----------+-----------+-----------+------------+-----------+-----------+
+2024/03/15 17:12:48 - patchstitcher - INFO - Epoch: [11/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.2545241117477417 - fine_loss: 1.2545241117477417
+2024/03/15 17:14:52 - patchstitcher - INFO - Epoch: [11/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.9477699398994446 - fine_loss: 0.9477699398994446
+2024/03/15 17:16:59 - patchstitcher - INFO - Epoch: [11/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.3806159496307373 - fine_loss: 1.3806159496307373
+2024/03/15 17:19:02 - patchstitcher - INFO - Epoch: [11/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.12031888961792 - fine_loss: 1.12031888961792
+2024/03/15 17:22:38 - patchstitcher - INFO - Epoch: [12/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.9633316993713379 - fine_loss: 0.9633316993713379
+2024/03/15 17:24:38 - patchstitcher - INFO - Epoch: [12/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.9473192691802979 - fine_loss: 0.9473192691802979
+2024/03/15 17:26:38 - patchstitcher - INFO - Epoch: [12/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.8891739845275879 - fine_loss: 0.8891739845275879
+2024/03/15 17:28:46 - patchstitcher - INFO - Epoch: [12/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.9305822849273682 - fine_loss: 0.9305822849273682
+2024/03/15 17:30:43 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  |  rmse_log |   silog   |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+| 0.9285209 | 0.9902661 | 0.9963124 | 0.0922186 | 1.4988106 | 0.0394503 | 0.1265562 | 11.929424 | 0.1792194 | 1.2142439 |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+2024/03/15 17:32:52 - patchstitcher - INFO - Epoch: [13/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.26497220993042 - fine_loss: 1.26497220993042
+2024/03/15 17:35:00 - patchstitcher - INFO - Epoch: [13/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.580217957496643 - fine_loss: 1.580217957496643
+2024/03/15 17:36:59 - patchstitcher - INFO - Epoch: [13/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.6395942568778992 - fine_loss: 0.6395942568778992
+2024/03/15 17:39:02 - patchstitcher - INFO - Epoch: [13/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.32594698667526245 - fine_loss: 0.32594698667526245
+2024/03/15 17:42:34 - patchstitcher - INFO - Epoch: [14/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.924031674861908 - fine_loss: 0.924031674861908
+2024/03/15 17:44:36 - patchstitcher - INFO - Epoch: [14/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.985018253326416 - fine_loss: 0.985018253326416
+2024/03/15 17:46:38 - patchstitcher - INFO - Epoch: [14/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.0442320108413696 - fine_loss: 1.0442320108413696
+2024/03/15 17:48:43 - patchstitcher - INFO - Epoch: [14/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.5068702101707458 - fine_loss: 0.5068702101707458
+2024/03/15 17:50:33 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  |  rmse_log |   silog    |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+| 0.9381619 | 0.9895476 | 0.9972216 | 0.0913334 | 1.5578288 | 0.0391697 | 0.1243245 | 11.1463653 | 0.1706981 | 1.1217431 |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+2024/03/15 17:52:46 - patchstitcher - INFO - Epoch: [15/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.1108862161636353 - fine_loss: 1.1108862161636353
+2024/03/15 17:54:52 - patchstitcher - INFO - Epoch: [15/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.9237959980964661 - fine_loss: 0.9237959980964661
+2024/03/15 17:56:56 - patchstitcher - INFO - Epoch: [15/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.5644421577453613 - fine_loss: 1.5644421577453613
+2024/03/15 17:58:54 - patchstitcher - INFO - Epoch: [15/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.7902756929397583 - fine_loss: 0.7902756929397583
+2024/03/15 18:02:26 - patchstitcher - INFO - Epoch: [16/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.966326117515564 - fine_loss: 0.966326117515564
+2024/03/15 18:04:32 - patchstitcher - INFO - Epoch: [16/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.9776898622512817 - fine_loss: 0.9776898622512817
+2024/03/15 18:06:33 - patchstitcher - INFO - Epoch: [16/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.6681317090988159 - fine_loss: 0.6681317090988159
+2024/03/15 18:08:34 - patchstitcher - INFO - Epoch: [16/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.80037522315979 - fine_loss: 0.80037522315979
+2024/03/15 18:10:20 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  |  rmse_log |   silog    |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+| 0.9538666 | 0.9917138 | 0.9972104 | 0.0811061 | 1.3823568 | 0.0351258 | 0.1140013 | 10.5376763 | 0.1382621 | 1.0577048 |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+2024/03/15 18:12:28 - patchstitcher - INFO - Epoch: [17/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.147787094116211 - fine_loss: 1.147787094116211
+2024/03/15 18:14:30 - patchstitcher - INFO - Epoch: [17/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.7300316691398621 - fine_loss: 0.7300316691398621
+2024/03/15 18:16:37 - patchstitcher - INFO - Epoch: [17/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.7750428318977356 - fine_loss: 0.7750428318977356
+2024/03/15 18:18:37 - patchstitcher - INFO - Epoch: [17/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.50600266456604 - fine_loss: 1.50600266456604
+2024/03/15 18:22:20 - patchstitcher - INFO - Epoch: [18/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.8911293745040894 - fine_loss: 0.8911293745040894
+2024/03/15 18:24:18 - patchstitcher - INFO - Epoch: [18/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.5605521202087402 - fine_loss: 0.5605521202087402
+2024/03/15 18:26:21 - patchstitcher - INFO - Epoch: [18/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.6763710975646973 - fine_loss: 1.6763710975646973
+2024/03/15 18:28:20 - patchstitcher - INFO - Epoch: [18/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.6500707864761353 - fine_loss: 0.6500707864761353
+2024/03/15 18:30:14 - patchstitcher - INFO - Evaluation Summary: 
++-----------+----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+|     a1    |    a2    |     a3    |  abs_rel  |    rmse   |   log_10  |  rmse_log |   silog    |   sq_rel  |    see    |
++-----------+----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+| 0.9562948 | 0.990871 | 0.9974688 | 0.0761721 | 1.3729287 | 0.0331131 | 0.1092103 | 10.1530306 | 0.1366973 | 1.0216396 |
++-----------+----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+2024/03/15 18:32:23 - patchstitcher - INFO - Epoch: [19/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.5755664110183716 - fine_loss: 0.5755664110183716
+2024/03/15 18:34:28 - patchstitcher - INFO - Epoch: [19/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.2044012546539307 - fine_loss: 1.2044012546539307
+2024/03/15 18:36:33 - patchstitcher - INFO - Epoch: [19/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.266536831855774 - fine_loss: 1.266536831855774
+2024/03/15 18:38:35 - patchstitcher - INFO - Epoch: [19/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.7211558818817139 - fine_loss: 0.7211558818817139
+2024/03/15 18:42:13 - patchstitcher - INFO - Epoch: [20/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.6136915683746338 - fine_loss: 0.6136915683746338
+2024/03/15 18:44:12 - patchstitcher - INFO - Epoch: [20/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.4747104048728943 - fine_loss: 0.4747104048728943
+2024/03/15 18:46:16 - patchstitcher - INFO - Epoch: [20/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.5850560069084167 - fine_loss: 0.5850560069084167
+2024/03/15 18:48:21 - patchstitcher - INFO - Epoch: [20/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.37204447388648987 - fine_loss: 0.37204447388648987
+2024/03/15 18:50:16 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+----------+-----------+-----------+-----------+-----------+-----------+-----------+----------+
+|     a1    |     a2    |    a3    |  abs_rel  |    rmse   |   log_10  |  rmse_log |   silog   |   sq_rel  |   see    |
++-----------+-----------+----------+-----------+-----------+-----------+-----------+-----------+-----------+----------+
+| 0.9645657 | 0.9920502 | 0.997654 | 0.0686085 | 1.2732928 | 0.0299144 | 0.1009926 | 9.6382305 | 0.1200509 | 0.993343 |
++-----------+-----------+----------+-----------+-----------+-----------+-----------+-----------+-----------+----------+
+2024/03/15 18:52:27 - patchstitcher - INFO - Epoch: [21/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.6047840714454651 - fine_loss: 0.6047840714454651
+2024/03/15 18:54:31 - patchstitcher - INFO - Epoch: [21/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.5551916360855103 - fine_loss: 0.5551916360855103
+2024/03/15 18:56:37 - patchstitcher - INFO - Epoch: [21/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.32560303807258606 - fine_loss: 0.32560303807258606
+2024/03/15 18:58:40 - patchstitcher - INFO - Epoch: [21/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.7431879043579102 - fine_loss: 1.7431879043579102
+2024/03/15 19:02:20 - patchstitcher - INFO - Epoch: [22/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.7936936020851135 - fine_loss: 0.7936936020851135
+2024/03/15 19:04:21 - patchstitcher - INFO - Epoch: [22/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.6791415214538574 - fine_loss: 0.6791415214538574
+2024/03/15 19:06:23 - patchstitcher - INFO - Epoch: [22/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.6265323758125305 - fine_loss: 0.6265323758125305
+2024/03/15 19:08:25 - patchstitcher - INFO - Epoch: [22/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.6945874691009521 - fine_loss: 0.6945874691009521
+2024/03/15 19:10:17 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  |  rmse_log |   silog   |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+| 0.9671118 | 0.9931541 | 0.9976758 | 0.0652155 | 1.2549019 | 0.0282474 | 0.0973396 | 9.2669667 | 0.1172386 | 0.9884787 |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+2024/03/15 19:12:25 - patchstitcher - INFO - Epoch: [23/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.2996392250061035 - fine_loss: 1.2996392250061035
+2024/03/15 19:14:26 - patchstitcher - INFO - Epoch: [23/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.674423098564148 - fine_loss: 0.674423098564148
+2024/03/15 19:16:29 - patchstitcher - INFO - Epoch: [23/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 2.0330402851104736 - fine_loss: 2.0330402851104736
+2024/03/15 19:18:34 - patchstitcher - INFO - Epoch: [23/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.1583242416381836 - fine_loss: 1.1583242416381836
+2024/03/15 19:22:12 - patchstitcher - INFO - Epoch: [24/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.8227792978286743 - fine_loss: 0.8227792978286743
+2024/03/15 19:24:12 - patchstitcher - INFO - Epoch: [24/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.6849284172058105 - fine_loss: 0.6849284172058105
+2024/03/15 19:26:14 - patchstitcher - INFO - Epoch: [24/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.5954287648200989 - fine_loss: 0.5954287648200989
+2024/03/15 19:28:20 - patchstitcher - INFO - Epoch: [24/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.38687634468078613 - fine_loss: 0.38687634468078613
+2024/03/15 19:30:07 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+----------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  |  rmse_log |  silog   |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+----------+-----------+-----------+
+| 0.9687062 | 0.9931654 | 0.9976169 | 0.0635503 | 1.2467909 | 0.0277027 | 0.0958232 | 9.191893 | 0.1155029 | 0.9803023 |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+----------+-----------+-----------+
+2024/03/15 19:30:07 - patchstitcher - INFO - Saving ckp, but use the inner get_save_dict fuction to get model_dict
+2024/03/15 19:30:07 - patchstitcher - INFO - For saving space. Would you like to save base model several times? :>
+2024/03/15 19:30:08 - patchstitcher - INFO - save checkpoint_24.pth at ./work_dir/depthanything_vitb_u4k/fine_pretrain

depthanything_vitb_u4k/fine_pretrain/checkpoint_24.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a2b0ca89e141a9a52626174d614584fccb47e140090a495cc5822803dac7018c
+size 1171453994

depthanything_vitb_u4k/fine_pretrain/config.py

@@ -0,0 +1,314 @@
+collect_input_args = [
+    'image_lr',
+    'crops_image_hr',
+    'depth_gt',
+    'crop_depths',
+    'bboxs',
+    'image_hr',
+]
+convert_syncbn = True
+debug = False
+env_cfg = dict(
+    cudnn_benchmark=True,
+    dist_cfg=dict(backend='nccl'),
+    mp_cfg=dict(mp_start_method='forkserver'))
+find_unused_parameters = True
+general_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        dataset_name='', gt_dir=None, rgb_image_dir='', type='ImageDataset'),
+    num_workers=2)
+launcher = 'pytorch'
+log_name = 'fine_pretrain'
+max_depth = 80
+min_depth = 0.001
+model = dict(
+    coarse_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        depth_anything=True,
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            392,
+            518,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='vitb',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource='local::./work_dir/DepthAnything_vitb.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='DA-ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    fine_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        depth_anything=True,
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            392,
+            518,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='vitb',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource='local::./work_dir/DepthAnything_vitb.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='DA-ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    max_depth=80,
+    min_depth=0.001,
+    patch_process_shape=(
+        392,
+        518,
+    ),
+    sigloss=dict(type='SILogLoss'),
+    target='fine',
+    type='BaselinePretrain')
+optim_wrapper = dict(
+    clip_grad=dict(max_norm=0.1, norm_type=2, type='norm'),
+    optimizer=dict(lr=4e-06, type='AdamW', weight_decay=0.01),
+    paramwise_cfg=dict(bypass_duplicate=True, custom_keys=dict()))
+param_scheduler = dict(
+    base_momentum=0.85,
+    cycle_momentum=True,
+    div_factor=1,
+    final_div_factor=10000,
+    max_momentum=0.95,
+    pct_start=0.5,
+    three_phase=False)
+project = 'patchfusion'
+resume = False
+tags = [
+    'fine',
+    'da',
+    'vitb',
+]
+test_in_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+test_out_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test_out.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+train_cfg = dict(
+    eval_start=0,
+    log_interval=100,
+    max_epochs=24,
+    save_checkpoint_interval=24,
+    train_log_img_interval=500,
+    val_interval=2,
+    val_log_img_interval=50,
+    val_type='epoch_base')
+train_dataloader = dict(
+    batch_size=4,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='train',
+        resize_mode='depth-anything',
+        split='./data/u4k/splits/train.txt',
+        transform_cfg=dict(
+            degree=1.0,
+            network_process_size=[
+                392,
+                518,
+            ],
+            random_crop=True,
+            random_crop_size=(
+                540,
+                960,
+            )),
+        type='UnrealStereo4kDataset'),
+    num_workers=4)
+val_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        resize_mode='depth-anything',
+        split='./data/u4k/splits/val.txt',
+        transform_cfg=dict(
+            degree=1.0,
+            network_process_size=[
+                392,
+                518,
+            ],
+            random_crop_size=(
+                540,
+                960,
+            )),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+work_dir = './work_dir/depthanything_vitb_u4k/fine_pretrain'
+zoe_depth_config = dict(
+    attractor_alpha=1000,
+    attractor_gamma=2,
+    attractor_kind='mean',
+    attractor_type='inv',
+    aug=True,
+    bin_centers_type='softplus',
+    bin_embedding_dim=128,
+    clip_grad=0.1,
+    dataset='nyu',
+    depth_anything=True,
+    distributed=True,
+    do_resize=False,
+    force_keep_ar=True,
+    freeze_midas_bn=True,
+    gpu='NULL',
+    img_size=[
+        392,
+        518,
+    ],
+    inverse_midas=False,
+    log_images_every=0.1,
+    max_depth=80,
+    max_temp=50.0,
+    max_translation=100,
+    memory_efficient=True,
+    midas_model_type='vitb',
+    min_depth=0.001,
+    min_temp=0.0212,
+    model='zoedepth',
+    n_attractors=[
+        16,
+        8,
+        4,
+        1,
+    ],
+    n_bins=64,
+    name='ZoeDepth',
+    notes='',
+    output_distribution='logbinomial',
+    prefetch=False,
+    pretrained_resource='local::./work_dir/DepthAnything_vitb.pt',
+    print_losses=False,
+    project='ZoeDepth',
+    random_crop=False,
+    random_translate=False,
+    root='.',
+    save_dir='',
+    shared_dict='NULL',
+    tags='',
+    train_midas=True,
+    translate_prob=0.2,
+    type='DA-ZoeDepth',
+    uid='NULL',
+    use_amp=False,
+    use_pretrained_midas=True,
+    use_shared_dict=False,
+    validate_every=0.25,
+    version_name='v1',
+    workers=16)

depthanything_vitb_u4k/patchfusion/checkpoint_16.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f104b19568f39e85783de4cd4ecf032ee24152e8daae929db458b111aef6ea20
+size 417857453

depthanything_vitb_u4k/patchfusion/config.py

@@ -0,0 +1,341 @@
+collect_input_args = [
+    'image_lr',
+    'crops_image_hr',
+    'depth_gt',
+    'crop_depths',
+    'bboxs',
+    'image_hr',
+]
+convert_syncbn = True
+debug = False
+env_cfg = dict(
+    cudnn_benchmark=True,
+    dist_cfg=dict(backend='nccl'),
+    mp_cfg=dict(mp_start_method='forkserver'))
+find_unused_parameters = True
+general_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        dataset_name='', gt_dir=None, rgb_image_dir='', type='ImageDataset'),
+    num_workers=2)
+launcher = 'pytorch'
+log_name = 'patchfusion'
+max_depth = 80
+min_depth = 0.001
+model = dict(
+    coarse_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        depth_anything=True,
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            392,
+            518,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='vitb',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource='local::./work_dir/DepthAnything_vitb.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='DA-ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    fine_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        depth_anything=True,
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            392,
+            518,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='vitb',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource='local::./work_dir/DepthAnything_vitb.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='DA-ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    guided_fusion=dict(
+        g2l=True,
+        in_channels=[
+            32,
+            128,
+            128,
+            128,
+            128,
+            128,
+        ],
+        n_channels=5,
+        num_patches=[
+            203056,
+            66304,
+            16576,
+            4144,
+            1036,
+            266,
+        ],
+        patch_process_shape=(
+            392,
+            518,
+        ),
+        type='GuidedFusionPatchFusion'),
+    max_depth=80,
+    min_depth=0.001,
+    patch_process_shape=(
+        392,
+        518,
+    ),
+    pretrain_model=[
+        './work_dir/depthanything_vitb_u4k/coarse_pretrain/checkpoint_24.pth',
+        './work_dir/depthanything_vitb_u4k/fine_pretrain/checkpoint_24.pth',
+    ],
+    sigloss=dict(type='SILogLoss'),
+    type='PatchFusion')
+optim_wrapper = dict(
+    clip_grad=dict(max_norm=0.1, norm_type=2, type='norm'),
+    optimizer=dict(lr=0.0001, type='AdamW', weight_decay=0.001),
+    paramwise_cfg=dict(bypass_duplicate=True, custom_keys=dict()))
+param_scheduler = dict(
+    base_momentum=0.85,
+    cycle_momentum=True,
+    div_factor=10,
+    final_div_factor=10000,
+    max_momentum=0.95,
+    pct_start=0.25,
+    three_phase=False)
+project = 'patchfusion'
+resume = False
+tags = [
+    'patchfusion',
+    'da',
+    'vitb',
+]
+test_in_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+test_out_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test_out.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+train_cfg = dict(
+    eval_start=0,
+    log_interval=100,
+    max_epochs=16,
+    save_checkpoint_interval=16,
+    train_log_img_interval=500,
+    val_interval=2,
+    val_log_img_interval=50,
+    val_type='epoch_base')
+train_dataloader = dict(
+    batch_size=4,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='train',
+        resize_mode='depth-anything',
+        split='./data/u4k/splits/train.txt',
+        transform_cfg=dict(
+            degree=1.0,
+            network_process_size=[
+                392,
+                518,
+            ],
+            random_crop=True,
+            random_crop_size=(
+                540,
+                960,
+            )),
+        type='UnrealStereo4kDataset'),
+    num_workers=4)
+val_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        resize_mode='depth-anything',
+        split='./data/u4k/splits/val.txt',
+        transform_cfg=dict(
+            degree=1.0,
+            network_process_size=[
+                392,
+                518,
+            ],
+            random_crop_size=(
+                540,
+                960,
+            )),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+work_dir = './work_dir/depthanything_vitb_u4k/patchfusion'
+zoe_depth_config = dict(
+    attractor_alpha=1000,
+    attractor_gamma=2,
+    attractor_kind='mean',
+    attractor_type='inv',
+    aug=True,
+    bin_centers_type='softplus',
+    bin_embedding_dim=128,
+    clip_grad=0.1,
+    dataset='nyu',
+    depth_anything=True,
+    distributed=True,
+    do_resize=False,
+    force_keep_ar=True,
+    freeze_midas_bn=True,
+    gpu='NULL',
+    img_size=[
+        392,
+        518,
+    ],
+    inverse_midas=False,
+    log_images_every=0.1,
+    max_depth=80,
+    max_temp=50.0,
+    max_translation=100,
+    memory_efficient=True,
+    midas_model_type='vitb',
+    min_depth=0.001,
+    min_temp=0.0212,
+    model='zoedepth',
+    n_attractors=[
+        16,
+        8,
+        4,
+        1,
+    ],
+    n_bins=64,
+    name='ZoeDepth',
+    notes='',
+    output_distribution='logbinomial',
+    prefetch=False,
+    pretrained_resource='local::./work_dir/DepthAnything_vitb.pt',
+    print_losses=False,
+    project='ZoeDepth',
+    random_crop=False,
+    random_translate=False,
+    root='.',
+    save_dir='',
+    shared_dict='NULL',
+    tags='',
+    train_midas=True,
+    translate_prob=0.2,
+    type='DA-ZoeDepth',
+    uid='NULL',
+    use_amp=False,
+    use_pretrained_midas=True,
+    use_shared_dict=False,
+    validate_every=0.25,
+    version_name='v1',
+    workers=16)

depthanything_vitl_u4k/coarse_pretrain/checkpoint_24.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7347385b649cf4a99cbd1cad579bdcdd51cab915bfd283031e55f7e718178f68
+size 4020717194

depthanything_vitl_u4k/coarse_pretrain/config.py

@@ -0,0 +1,310 @@
+collect_input_args = [
+    'image_lr',
+    'crops_image_hr',
+    'depth_gt',
+    'crop_depths',
+    'bboxs',
+    'image_hr',
+]
+convert_syncbn = True
+debug = False
+env_cfg = dict(
+    cudnn_benchmark=True,
+    dist_cfg=dict(backend='nccl'),
+    mp_cfg=dict(mp_start_method='forkserver'))
+find_unused_parameters = True
+general_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        dataset_name='', gt_dir=None, rgb_image_dir='', type='ImageDataset'),
+    num_workers=2)
+launcher = 'pytorch'
+log_name = 'coarse_pretrain'
+max_depth = 80
+min_depth = 0.001
+model = dict(
+    coarse_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        depth_anything=True,
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            392,
+            518,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='vitl',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource='local::./work_dir/DepthAnything_vitl.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='DA-ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    fine_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        depth_anything=True,
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            392,
+            518,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='vitl',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource='local::./work_dir/DepthAnything_vitl.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='DA-ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    max_depth=80,
+    min_depth=0.001,
+    sigloss=dict(type='SILogLoss'),
+    target='coarse',
+    type='BaselinePretrain')
+optim_wrapper = dict(
+    clip_grad=dict(max_norm=0.1, norm_type=2, type='norm'),
+    optimizer=dict(lr=4e-06, type='AdamW', weight_decay=0.01),
+    paramwise_cfg=dict(bypass_duplicate=True, custom_keys=dict()))
+param_scheduler = dict(
+    base_momentum=0.85,
+    cycle_momentum=True,
+    div_factor=1,
+    final_div_factor=10000,
+    max_momentum=0.95,
+    pct_start=0.5,
+    three_phase=False)
+project = 'patchfusion'
+resume = False
+tags = [
+    'coarse',
+    'da',
+    'vitl',
+]
+test_in_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+test_out_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test_out.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+train_cfg = dict(
+    eval_start=0,
+    log_interval=100,
+    max_epochs=24,
+    save_checkpoint_interval=24,
+    train_log_img_interval=500,
+    val_interval=2,
+    val_log_img_interval=50,
+    val_type='epoch_base')
+train_dataloader = dict(
+    batch_size=4,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='train',
+        resize_mode='depth-anything',
+        split='./data/u4k/splits/train.txt',
+        transform_cfg=dict(
+            degree=1.0,
+            network_process_size=[
+                392,
+                518,
+            ],
+            random_crop=True,
+            random_crop_size=(
+                540,
+                960,
+            )),
+        type='UnrealStereo4kDataset'),
+    num_workers=4)
+val_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        resize_mode='depth-anything',
+        split='./data/u4k/splits/val.txt',
+        transform_cfg=dict(
+            degree=1.0,
+            network_process_size=[
+                392,
+                518,
+            ],
+            random_crop_size=(
+                540,
+                960,
+            )),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+work_dir = './work_dir/depthanything_vitl_u4k/coarse_pretrain'
+zoe_depth_config = dict(
+    attractor_alpha=1000,
+    attractor_gamma=2,
+    attractor_kind='mean',
+    attractor_type='inv',
+    aug=True,
+    bin_centers_type='softplus',
+    bin_embedding_dim=128,
+    clip_grad=0.1,
+    dataset='nyu',
+    depth_anything=True,
+    distributed=True,
+    do_resize=False,
+    force_keep_ar=True,
+    freeze_midas_bn=True,
+    gpu='NULL',
+    img_size=[
+        392,
+        518,
+    ],
+    inverse_midas=False,
+    log_images_every=0.1,
+    max_depth=80,
+    max_temp=50.0,
+    max_translation=100,
+    memory_efficient=True,
+    midas_model_type='vitl',
+    min_depth=0.001,
+    min_temp=0.0212,
+    model='zoedepth',
+    n_attractors=[
+        16,
+        8,
+        4,
+        1,
+    ],
+    n_bins=64,
+    name='ZoeDepth',
+    notes='',
+    output_distribution='logbinomial',
+    prefetch=False,
+    pretrained_resource='local::./work_dir/DepthAnything_vitl.pt',
+    print_losses=False,
+    project='ZoeDepth',
+    random_crop=False,
+    random_translate=False,
+    root='.',
+    save_dir='',
+    shared_dict='NULL',
+    tags='',
+    train_midas=True,
+    translate_prob=0.2,
+    type='DA-ZoeDepth',
+    uid='NULL',
+    use_amp=False,
+    use_pretrained_midas=True,
+    use_shared_dict=False,
+    validate_every=0.25,
+    version_name='v1',
+    workers=16)

depthanything_vitl_u4k/fine_pretrain/checkpoint_24.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:37181232060bc2b0fd663cf3fc008dda37b262f680a915689f6e55f072648fc7
+size 4020717194

depthanything_vitl_u4k/fine_pretrain/config.py

@@ -0,0 +1,314 @@
+collect_input_args = [
+    'image_lr',
+    'crops_image_hr',
+    'depth_gt',
+    'crop_depths',
+    'bboxs',
+    'image_hr',
+]
+convert_syncbn = True
+debug = False
+env_cfg = dict(
+    cudnn_benchmark=True,
+    dist_cfg=dict(backend='nccl'),
+    mp_cfg=dict(mp_start_method='forkserver'))
+find_unused_parameters = True
+general_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        dataset_name='', gt_dir=None, rgb_image_dir='', type='ImageDataset'),
+    num_workers=2)
+launcher = 'pytorch'
+log_name = 'fine_pretrain'
+max_depth = 80
+min_depth = 0.001
+model = dict(
+    coarse_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        depth_anything=True,
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            392,
+            518,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='vitl',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource='local::./work_dir/DepthAnything_vitl.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='DA-ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    fine_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        depth_anything=True,
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            392,
+            518,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='vitl',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource='local::./work_dir/DepthAnything_vitl.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='DA-ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    max_depth=80,
+    min_depth=0.001,
+    patch_process_shape=(
+        392,
+        518,
+    ),
+    sigloss=dict(type='SILogLoss'),
+    target='fine',
+    type='BaselinePretrain')
+optim_wrapper = dict(
+    clip_grad=dict(max_norm=0.1, norm_type=2, type='norm'),
+    optimizer=dict(lr=4e-06, type='AdamW', weight_decay=0.01),
+    paramwise_cfg=dict(bypass_duplicate=True, custom_keys=dict()))
+param_scheduler = dict(
+    base_momentum=0.85,
+    cycle_momentum=True,
+    div_factor=1,
+    final_div_factor=10000,
+    max_momentum=0.95,
+    pct_start=0.5,
+    three_phase=False)
+project = 'patchfusion'
+resume = False
+tags = [
+    'fine',
+    'da',
+    'vitl',
+]
+test_in_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+test_out_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test_out.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+train_cfg = dict(
+    eval_start=0,
+    log_interval=100,
+    max_epochs=24,
+    save_checkpoint_interval=24,
+    train_log_img_interval=500,
+    val_interval=2,
+    val_log_img_interval=50,
+    val_type='epoch_base')
+train_dataloader = dict(
+    batch_size=4,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='train',
+        resize_mode='depth-anything',
+        split='./data/u4k/splits/train.txt',
+        transform_cfg=dict(
+            degree=1.0,
+            network_process_size=[
+                392,
+                518,
+            ],
+            random_crop=True,
+            random_crop_size=(
+                540,
+                960,
+            )),
+        type='UnrealStereo4kDataset'),
+    num_workers=4)
+val_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        resize_mode='depth-anything',
+        split='./data/u4k/splits/val.txt',
+        transform_cfg=dict(
+            degree=1.0,
+            network_process_size=[
+                392,
+                518,
+            ],
+            random_crop_size=(
+                540,
+                960,
+            )),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+work_dir = './work_dir/depthanything_vitl_u4k/fine_pretrain'
+zoe_depth_config = dict(
+    attractor_alpha=1000,
+    attractor_gamma=2,
+    attractor_kind='mean',
+    attractor_type='inv',
+    aug=True,
+    bin_centers_type='softplus',
+    bin_embedding_dim=128,
+    clip_grad=0.1,
+    dataset='nyu',
+    depth_anything=True,
+    distributed=True,
+    do_resize=False,
+    force_keep_ar=True,
+    freeze_midas_bn=True,
+    gpu='NULL',
+    img_size=[
+        392,
+        518,
+    ],
+    inverse_midas=False,
+    log_images_every=0.1,
+    max_depth=80,
+    max_temp=50.0,
+    max_translation=100,
+    memory_efficient=True,
+    midas_model_type='vitl',
+    min_depth=0.001,
+    min_temp=0.0212,
+    model='zoedepth',
+    n_attractors=[
+        16,
+        8,
+        4,
+        1,
+    ],
+    n_bins=64,
+    name='ZoeDepth',
+    notes='',
+    output_distribution='logbinomial',
+    prefetch=False,
+    pretrained_resource='local::./work_dir/DepthAnything_vitl.pt',
+    print_losses=False,
+    project='ZoeDepth',
+    random_crop=False,
+    random_translate=False,
+    root='.',
+    save_dir='',
+    shared_dict='NULL',
+    tags='',
+    train_midas=True,
+    translate_prob=0.2,
+    type='DA-ZoeDepth',
+    uid='NULL',
+    use_amp=False,
+    use_pretrained_midas=True,
+    use_shared_dict=False,
+    validate_every=0.25,
+    version_name='v1',
+    workers=16)

depthanything_vitl_u4k/patchfusion/checkpoint_16.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:79e530dd2ad7587b21b2b778b60ba0c459621969ed8b015b96987124c0747e10
+size 1128275629

depthanything_vitl_u4k/patchfusion/config.py

@@ -0,0 +1,347 @@
+collect_input_args = [
+    'image_lr',
+    'crops_image_hr',
+    'depth_gt',
+    'crop_depths',
+    'bboxs',
+    'image_hr',
+]
+convert_syncbn = True
+debug = True
+env_cfg = dict(
+    cudnn_benchmark=True,
+    dist_cfg=dict(backend='nccl'),
+    mp_cfg=dict(mp_start_method='forkserver'))
+find_unused_parameters = True
+general_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        dataset_name='',
+        gt_dir=None,
+        network_process_size=(
+            392,
+            518,
+        ),
+        resize_mode='depth-anything',
+        rgb_image_dir='',
+        type='ImageDataset'),
+    num_workers=2)
+launcher = 'pytorch'
+log_name = 'patchfusion'
+max_depth = 80
+min_depth = 0.001
+model = dict(
+    coarse_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        depth_anything=True,
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            392,
+            518,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='vitl',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource='local::./work_dir/DepthAnything_vitl.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='DA-ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    fine_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        depth_anything=True,
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            392,
+            518,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='vitl',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource='local::./work_dir/DepthAnything_vitl.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='DA-ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    guided_fusion=dict(
+        g2l=True,
+        in_channels=[
+            32,
+            256,
+            256,
+            256,
+            256,
+            256,
+        ],
+        n_channels=5,
+        num_patches=[
+            203056,
+            66304,
+            16576,
+            4144,
+            1036,
+            266,
+        ],
+        patch_process_shape=(
+            392,
+            518,
+        ),
+        type='GuidedFusionPatchFusion'),
+    max_depth=80,
+    min_depth=0.001,
+    patch_process_shape=(
+        392,
+        518,
+    ),
+    pretrain_model=[
+        './work_dir/depthanything_vitl_u4k/coarse_pretrain/checkpoint_24.pth',
+        './work_dir/depthanything_vitl_u4k/fine_pretrain/checkpoint_24.pth',
+    ],
+    sigloss=dict(type='SILogLoss'),
+    type='PatchFusion')
+optim_wrapper = dict(
+    clip_grad=dict(max_norm=0.1, norm_type=2, type='norm'),
+    optimizer=dict(lr=0.0001, type='AdamW', weight_decay=0.001),
+    paramwise_cfg=dict(bypass_duplicate=True, custom_keys=dict()))
+param_scheduler = dict(
+    base_momentum=0.85,
+    cycle_momentum=True,
+    div_factor=10,
+    final_div_factor=10000,
+    max_momentum=0.95,
+    pct_start=0.25,
+    three_phase=False)
+project = 'patchfusion'
+resume = False
+tags = [
+    'patchfusion',
+    'da',
+    'vitl',
+]
+test_in_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+test_out_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test_out.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+train_cfg = dict(
+    eval_start=0,
+    log_interval=100,
+    max_epochs=16,
+    save_checkpoint_interval=16,
+    train_log_img_interval=500,
+    val_interval=2,
+    val_log_img_interval=50,
+    val_type='epoch_base')
+train_dataloader = dict(
+    batch_size=4,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='train',
+        resize_mode='depth-anything',
+        split='./data/u4k/splits/train.txt',
+        transform_cfg=dict(
+            degree=1.0,
+            network_process_size=[
+                392,
+                518,
+            ],
+            random_crop=True,
+            random_crop_size=(
+                540,
+                960,
+            )),
+        type='UnrealStereo4kDataset'),
+    num_workers=4)
+val_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        resize_mode='depth-anything',
+        split='./data/u4k/splits/val.txt',
+        transform_cfg=dict(
+            network_process_size=[
+                392,
+                518,
+            ], random_crop_size=(
+                540,
+                960,
+            )),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+work_dir = './work_dir/depthanything_vitl_u4k/patchfusion'
+zoe_depth_config = dict(
+    attractor_alpha=1000,
+    attractor_gamma=2,
+    attractor_kind='mean',
+    attractor_type='inv',
+    aug=True,
+    bin_centers_type='softplus',
+    bin_embedding_dim=128,
+    clip_grad=0.1,
+    dataset='nyu',
+    depth_anything=True,
+    distributed=True,
+    do_resize=False,
+    force_keep_ar=True,
+    freeze_midas_bn=True,
+    gpu='NULL',
+    img_size=[
+        392,
+        518,
+    ],
+    inverse_midas=False,
+    log_images_every=0.1,
+    max_depth=80,
+    max_temp=50.0,
+    max_translation=100,
+    memory_efficient=True,
+    midas_model_type='vitl',
+    min_depth=0.001,
+    min_temp=0.0212,
+    model='zoedepth',
+    n_attractors=[
+        16,
+        8,
+        4,
+        1,
+    ],
+    n_bins=64,
+    name='ZoeDepth',
+    notes='',
+    output_distribution='logbinomial',
+    prefetch=False,
+    pretrained_resource='local::./work_dir/DepthAnything_vitl.pt',
+    print_losses=False,
+    project='ZoeDepth',
+    random_crop=False,
+    random_translate=False,
+    root='.',
+    save_dir='',
+    shared_dict='NULL',
+    tags='',
+    train_midas=True,
+    translate_prob=0.2,
+    type='DA-ZoeDepth',
+    uid='NULL',
+    use_amp=False,
+    use_pretrained_midas=True,
+    use_shared_dict=False,
+    validate_every=0.25,
+    version_name='v1',
+    workers=16)

depthanything_vits_u4k/coarse_pretrain/20240315_002030.log

@@ -0,0 +1,1024 @@
+2024/03/15 00:20:41 - patchstitcher - INFO - 
+------------------------------------------------------------
+System environment:
+    sys.platform: linux
+    Python: 3.8.18 | packaged by conda-forge | (default, Oct 10 2023, 15:44:36) [GCC 12.3.0]
+    CUDA available: True
+    numpy_random_seed: 621
+    GPU 0,1,2,3: NVIDIA A100-SXM4-80GB
+    CUDA_HOME: /sw/rl9g/cuda/11.8/rl9_binary
+    NVCC: Cuda compilation tools, release 11.8, V11.8.89
+    GCC: gcc (GCC) 11.3.1 20220421 (Red Hat 11.3.1-2)
+    PyTorch: 2.1.2
+    PyTorch compiling details: PyTorch built with:
+  - GCC 9.3
+  - C++ Version: 201703
+  - Intel(R) oneAPI Math Kernel Library Version 2022.1-Product Build 20220311 for Intel(R) 64 architecture applications
+  - Intel(R) MKL-DNN v3.1.1 (Git Hash 64f6bcbcbab628e96f33a62c3e975f8535a7bde4)
+  - OpenMP 201511 (a.k.a. OpenMP 4.5)
+  - LAPACK is enabled (usually provided by MKL)
+  - NNPACK is enabled
+  - CPU capability usage: AVX2
+  - CUDA Runtime 11.8
+  - NVCC architecture flags: -gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_61,code=sm_61;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_80,code=sm_80;-gencode;arch=compute_86,code=sm_86;-gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_90,code=sm_90;-gencode;arch=compute_37,code=compute_37
+  - CuDNN 8.7
+  - Magma 2.6.1
+  - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=11.8, CUDNN_VERSION=8.7.0, CXX_COMPILER=/opt/rh/devtoolset-9/root/usr/bin/c++, CXX_FLAGS= -D_GLIBCXX_USE_CXX11_ABI=0 -fabi-version=11 -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOROCTRACER -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -O2 -fPIC -Wall -Wextra -Werror=return-type -Werror=non-virtual-dtor -Werror=bool-operation -Wnarrowing -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wno-unused-parameter -Wno-unused-function -Wno-unused-result -Wno-strict-overflow -Wno-strict-aliasing -Wno-stringop-overflow -Wno-psabi -Wno-error=pedantic -Wno-error=old-style-cast -Wno-invalid-partial-specialization -Wno-unused-private-field -Wno-aligned-allocation-unavailable -Wno-missing-braces -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Werror=cast-function-type -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_DISABLE_GPU_ASSERTS=ON, TORCH_VERSION=2.1.2, USE_CUDA=ON, USE_CUDNN=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=ON, USE_NNPACK=ON, USE_OPENMP=ON, USE_ROCM=OFF, 
+
+    TorchVision: 0.16.2
+    OpenCV: 4.8.1
+    MMEngine: 0.10.2
+
+Runtime environment:
+    cudnn_benchmark: True
+    mp_cfg: {'mp_start_method': 'forkserver'}
+    dist_cfg: {'backend': 'nccl'}
+    seed: 621
+    Distributed launcher: pytorch
+    Distributed training: True
+    GPU number: 4
+------------------------------------------------------------
+
+2024/03/15 00:20:41 - patchstitcher - INFO - Config:
+collect_input_args = [
+    'image_lr',
+    'crops_image_hr',
+    'depth_gt',
+    'crop_depths',
+    'bboxs',
+    'image_hr',
+]
+convert_syncbn = True
+debug = False
+env_cfg = dict(
+    cudnn_benchmark=True,
+    dist_cfg=dict(backend='nccl'),
+    mp_cfg=dict(mp_start_method='forkserver'))
+find_unused_parameters = True
+general_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        dataset_name='', gt_dir=None, rgb_image_dir='', type='ImageDataset'),
+    num_workers=2)
+launcher = 'pytorch'
+log_name = 'coarse_pretrain'
+max_depth = 80
+min_depth = 0.001
+model = dict(
+    coarse_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        depth_anything=True,
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            392,
+            518,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='vits',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource='local::./work_dir/DepthAnything_vits.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='DA-ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    fine_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        depth_anything=True,
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            392,
+            518,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='vits',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource='local::./work_dir/DepthAnything_vits.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='DA-ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    max_depth=80,
+    min_depth=0.001,
+    sigloss=dict(type='SILogLoss'),
+    target='coarse',
+    type='BaselinePretrain')
+optim_wrapper = dict(
+    clip_grad=dict(max_norm=0.1, norm_type=2, type='norm'),
+    optimizer=dict(lr=4e-06, type='AdamW', weight_decay=0.01),
+    paramwise_cfg=dict(bypass_duplicate=True, custom_keys=dict()))
+param_scheduler = dict(
+    base_momentum=0.85,
+    cycle_momentum=True,
+    div_factor=1,
+    final_div_factor=10000,
+    max_momentum=0.95,
+    pct_start=0.5,
+    three_phase=False)
+project = 'patchfusion'
+tags = [
+    'coarse',
+    'da',
+    'vits',
+]
+test_in_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+test_out_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test_out.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+train_cfg = dict(
+    eval_start=0,
+    log_interval=100,
+    max_epochs=24,
+    save_checkpoint_interval=24,
+    train_log_img_interval=100,
+    val_interval=2,
+    val_log_img_interval=50,
+    val_type='epoch_base')
+train_dataloader = dict(
+    batch_size=4,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='train',
+        resize_mode='depth-anything',
+        split='./data/u4k/splits/train.txt',
+        transform_cfg=dict(
+            degree=1.0, network_process_size=[
+                392,
+                518,
+            ], random_crop=True),
+        type='UnrealStereo4kDataset'),
+    num_workers=4)
+val_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        resize_mode='depth-anything',
+        split='./data/u4k/splits/val.txt',
+        transform_cfg=dict(degree=1.0, network_process_size=[
+            392,
+            518,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+work_dir = './work_dir/depthanything_vits_u4k/coarse_pretrain'
+zoe_depth_config = dict(
+    attractor_alpha=1000,
+    attractor_gamma=2,
+    attractor_kind='mean',
+    attractor_type='inv',
+    aug=True,
+    bin_centers_type='softplus',
+    bin_embedding_dim=128,
+    clip_grad=0.1,
+    dataset='nyu',
+    depth_anything=True,
+    distributed=True,
+    do_resize=False,
+    force_keep_ar=True,
+    freeze_midas_bn=True,
+    gpu='NULL',
+    img_size=[
+        392,
+        518,
+    ],
+    inverse_midas=False,
+    log_images_every=0.1,
+    max_depth=80,
+    max_temp=50.0,
+    max_translation=100,
+    memory_efficient=True,
+    midas_model_type='vits',
+    min_depth=0.001,
+    min_temp=0.0212,
+    model='zoedepth',
+    n_attractors=[
+        16,
+        8,
+        4,
+        1,
+    ],
+    n_bins=64,
+    name='ZoeDepth',
+    notes='',
+    output_distribution='logbinomial',
+    prefetch=False,
+    pretrained_resource='local::./work_dir/DepthAnything_vits.pt',
+    print_losses=False,
+    project='ZoeDepth',
+    random_crop=False,
+    random_translate=False,
+    root='.',
+    save_dir='',
+    shared_dict='NULL',
+    tags='',
+    train_midas=True,
+    translate_prob=0.2,
+    type='DA-ZoeDepth',
+    uid='NULL',
+    use_amp=False,
+    use_pretrained_midas=True,
+    use_shared_dict=False,
+    validate_every=0.25,
+    version_name='v1',
+    workers=16)
+
+2024/03/15 00:20:41 - patchstitcher - INFO - Loading deepnet from local::./work_dir/DepthAnything_vits.pt
+2024/03/15 00:20:41 - patchstitcher - INFO - Current zoedepth.core.prep.resizer is <class 'torch.nn.modules.linear.Identity'>
+2024/03/15 00:20:42 - patchstitcher - INFO - DistributedDataParallel(
+  (module): BaselinePretrain(
+    (coarse_branch): ZoeDepth(
+      (core): DepthAnythingCore(
+        (core): DPT_DINOv2(
+          (pretrained): DinoVisionTransformer(
+            (patch_embed): PatchEmbed(
+              (proj): Conv2d(3, 384, kernel_size=(14, 14), stride=(14, 14))
+              (norm): Identity()
+            )
+            (blocks): ModuleList(
+              (0-11): 12 x NestedTensorBlock(
+                (norm1): LayerNorm((384,), eps=1e-06, elementwise_affine=True)
+                (attn): MemEffAttention(
+                  (qkv): Linear(in_features=384, out_features=1152, bias=True)
+                  (attn_drop): Dropout(p=0.0, inplace=False)
+                  (proj): Linear(in_features=384, out_features=384, bias=True)
+                  (proj_drop): Dropout(p=0.0, inplace=False)
+                )
+                (ls1): LayerScale()
+                (drop_path1): Identity()
+                (norm2): LayerNorm((384,), eps=1e-06, elementwise_affine=True)
+                (mlp): Mlp(
+                  (fc1): Linear(in_features=384, out_features=1536, bias=True)
+                  (act): GELU(approximate='none')
+                  (fc2): Linear(in_features=1536, out_features=384, bias=True)
+                  (drop): Dropout(p=0.0, inplace=False)
+                )
+                (ls2): LayerScale()
+                (drop_path2): Identity()
+              )
+            )
+            (norm): LayerNorm((384,), eps=1e-06, elementwise_affine=True)
+            (head): Identity()
+          )
+          (depth_head): DPTHead(
+            (projects): ModuleList(
+              (0): Conv2d(384, 48, kernel_size=(1, 1), stride=(1, 1))
+              (1): Conv2d(384, 96, kernel_size=(1, 1), stride=(1, 1))
+              (2): Conv2d(384, 192, kernel_size=(1, 1), stride=(1, 1))
+              (3): Conv2d(384, 384, kernel_size=(1, 1), stride=(1, 1))
+            )
+            (resize_layers): ModuleList(
+              (0): ConvTranspose2d(48, 48, kernel_size=(4, 4), stride=(4, 4))
+              (1): ConvTranspose2d(96, 96, kernel_size=(2, 2), stride=(2, 2))
+              (2): Identity()
+              (3): Conv2d(384, 384, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1))
+            )
+            (scratch): Module(
+              (layer1_rn): Conv2d(48, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
+              (layer2_rn): Conv2d(96, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
+              (layer3_rn): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
+              (layer4_rn): Conv2d(384, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
+              (refinenet1): FeatureFusionBlock(
+                (out_conv): Conv2d(64, 64, kernel_size=(1, 1), stride=(1, 1))
+                (resConfUnit1): ResidualConvUnit(
+                  (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (resConfUnit2): ResidualConvUnit(
+                  (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (skip_add): FloatFunctional(
+                  (activation_post_process): Identity()
+                )
+              )
+              (refinenet2): FeatureFusionBlock(
+                (out_conv): Conv2d(64, 64, kernel_size=(1, 1), stride=(1, 1))
+                (resConfUnit1): ResidualConvUnit(
+                  (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (resConfUnit2): ResidualConvUnit(
+                  (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (skip_add): FloatFunctional(
+                  (activation_post_process): Identity()
+                )
+              )
+              (refinenet3): FeatureFusionBlock(
+                (out_conv): Conv2d(64, 64, kernel_size=(1, 1), stride=(1, 1))
+                (resConfUnit1): ResidualConvUnit(
+                  (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (resConfUnit2): ResidualConvUnit(
+                  (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (skip_add): FloatFunctional(
+                  (activation_post_process): Identity()
+                )
+              )
+              (refinenet4): FeatureFusionBlock(
+                (out_conv): Conv2d(64, 64, kernel_size=(1, 1), stride=(1, 1))
+                (resConfUnit1): ResidualConvUnit(
+                  (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (resConfUnit2): ResidualConvUnit(
+                  (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (skip_add): FloatFunctional(
+                  (activation_post_process): Identity()
+                )
+              )
+              (output_conv1): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+              (output_conv2): Sequential(
+                (0): Conv2d(32, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                (1): ReLU(inplace=True)
+                (2): Conv2d(32, 1, kernel_size=(1, 1), stride=(1, 1))
+                (3): ReLU(inplace=True)
+                (4): Identity()
+              )
+            )
+          )
+        )
+      )
+      (conv2): Conv2d(64, 64, kernel_size=(1, 1), stride=(1, 1))
+      (seed_bin_regressor): SeedBinRegressorUnnormed(
+        (_net): Sequential(
+          (0): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1))
+          (1): ReLU(inplace=True)
+          (2): Conv2d(256, 64, kernel_size=(1, 1), stride=(1, 1))
+          (3): Softplus(beta=1, threshold=20)
+        )
+      )
+      (seed_projector): Projector(
+        (_net): Sequential(
+          (0): Conv2d(64, 128, kernel_size=(1, 1), stride=(1, 1))
+          (1): ReLU(inplace=True)
+          (2): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+        )
+      )
+      (projectors): ModuleList(
+        (0-3): 4 x Projector(
+          (_net): Sequential(
+            (0): Conv2d(64, 128, kernel_size=(1, 1), stride=(1, 1))
+            (1): ReLU(inplace=True)
+            (2): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+          )
+        )
+      )
+      (attractors): ModuleList(
+        (0): AttractorLayerUnnormed(
+          (_net): Sequential(
+            (0): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+            (1): ReLU(inplace=True)
+            (2): Conv2d(128, 16, kernel_size=(1, 1), stride=(1, 1))
+            (3): Softplus(beta=1, threshold=20)
+          )
+        )
+        (1): AttractorLayerUnnormed(
+          (_net): Sequential(
+            (0): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+            (1): ReLU(inplace=True)
+            (2): Conv2d(128, 8, kernel_size=(1, 1), stride=(1, 1))
+            (3): Softplus(beta=1, threshold=20)
+          )
+        )
+        (2): AttractorLayerUnnormed(
+          (_net): Sequential(
+            (0): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+            (1): ReLU(inplace=True)
+            (2): Conv2d(128, 4, kernel_size=(1, 1), stride=(1, 1))
+            (3): Softplus(beta=1, threshold=20)
+          )
+        )
+        (3): AttractorLayerUnnormed(
+          (_net): Sequential(
+            (0): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+            (1): ReLU(inplace=True)
+            (2): Conv2d(128, 1, kernel_size=(1, 1), stride=(1, 1))
+            (3): Softplus(beta=1, threshold=20)
+          )
+        )
+      )
+      (conditional_log_binomial): ConditionalLogBinomial(
+        (log_binomial_transform): LogBinomial()
+        (mlp): Sequential(
+          (0): Conv2d(161, 80, kernel_size=(1, 1), stride=(1, 1))
+          (1): GELU(approximate='none')
+          (2): Conv2d(80, 4, kernel_size=(1, 1), stride=(1, 1))
+          (3): Softplus(beta=1, threshold=20)
+        )
+      )
+    )
+    (sigloss): SILogLoss()
+  )
+)
+2024/03/15 00:20:47 - patchstitcher - INFO - successfully init trainer
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.cls_token
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.pos_embed
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.mask_token
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.patch_embed.proj.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.patch_embed.proj.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.0.norm1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.0.norm1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.0.attn.qkv.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.0.attn.qkv.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.0.attn.proj.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.0.attn.proj.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.0.ls1.gamma
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.0.norm2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.0.norm2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.0.mlp.fc1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.0.mlp.fc1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.0.mlp.fc2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.0.mlp.fc2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.0.ls2.gamma
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.1.norm1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.1.norm1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.1.attn.qkv.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.1.attn.qkv.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.1.attn.proj.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.1.attn.proj.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.1.ls1.gamma
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.1.norm2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.1.norm2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.1.mlp.fc1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.1.mlp.fc1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.1.mlp.fc2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.1.mlp.fc2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.1.ls2.gamma
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.2.norm1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.2.norm1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.2.attn.qkv.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.2.attn.qkv.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.2.attn.proj.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.2.attn.proj.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.2.ls1.gamma
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.2.norm2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.2.norm2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.2.mlp.fc1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.2.mlp.fc1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.2.mlp.fc2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.2.mlp.fc2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.2.ls2.gamma
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.3.norm1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.3.norm1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.3.attn.qkv.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.3.attn.qkv.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.3.attn.proj.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.3.attn.proj.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.3.ls1.gamma
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.3.norm2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.3.norm2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.3.mlp.fc1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.3.mlp.fc1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.3.mlp.fc2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.3.mlp.fc2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.3.ls2.gamma
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.4.norm1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.4.norm1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.4.attn.qkv.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.4.attn.qkv.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.4.attn.proj.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.4.attn.proj.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.4.ls1.gamma
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.4.norm2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.4.norm2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.4.mlp.fc1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.4.mlp.fc1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.4.mlp.fc2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.4.mlp.fc2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.4.ls2.gamma
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.5.norm1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.5.norm1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.5.attn.qkv.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.5.attn.qkv.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.5.attn.proj.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.5.attn.proj.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.5.ls1.gamma
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.5.norm2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.5.norm2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.5.mlp.fc1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.5.mlp.fc1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.5.mlp.fc2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.5.mlp.fc2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.5.ls2.gamma
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.6.norm1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.6.norm1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.6.attn.qkv.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.6.attn.qkv.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.6.attn.proj.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.6.attn.proj.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.6.ls1.gamma
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.6.norm2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.6.norm2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.6.mlp.fc1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.6.mlp.fc1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.6.mlp.fc2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.6.mlp.fc2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.6.ls2.gamma
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.7.norm1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.7.norm1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.7.attn.qkv.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.7.attn.qkv.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.7.attn.proj.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.7.attn.proj.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.7.ls1.gamma
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.7.norm2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.7.norm2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.7.mlp.fc1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.7.mlp.fc1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.7.mlp.fc2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.7.mlp.fc2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.7.ls2.gamma
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.8.norm1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.8.norm1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.8.attn.qkv.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.8.attn.qkv.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.8.attn.proj.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.8.attn.proj.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.8.ls1.gamma
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.8.norm2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.8.norm2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.8.mlp.fc1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.8.mlp.fc1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.8.mlp.fc2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.8.mlp.fc2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.8.ls2.gamma
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.9.norm1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.9.norm1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.9.attn.qkv.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.9.attn.qkv.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.9.attn.proj.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.9.attn.proj.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.9.ls1.gamma
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.9.norm2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.9.norm2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.9.mlp.fc1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.9.mlp.fc1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.9.mlp.fc2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.9.mlp.fc2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.9.ls2.gamma
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.10.norm1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.10.norm1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.10.attn.qkv.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.10.attn.qkv.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.10.attn.proj.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.10.attn.proj.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.10.ls1.gamma
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.10.norm2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.10.norm2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.10.mlp.fc1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.10.mlp.fc1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.10.mlp.fc2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.10.mlp.fc2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.10.ls2.gamma
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.11.norm1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.11.norm1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.11.attn.qkv.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.11.attn.qkv.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.11.attn.proj.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.11.attn.proj.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.11.ls1.gamma
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.11.norm2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.11.norm2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.11.mlp.fc1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.11.mlp.fc1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.11.mlp.fc2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.11.mlp.fc2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.blocks.11.ls2.gamma
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.norm.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.pretrained.norm.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.projects.0.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.projects.0.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.projects.1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.projects.1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.projects.2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.projects.2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.projects.3.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.projects.3.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.resize_layers.0.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.resize_layers.0.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.resize_layers.1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.resize_layers.1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.resize_layers.3.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.resize_layers.3.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.layer1_rn.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.layer2_rn.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.layer3_rn.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.layer4_rn.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet1.out_conv.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet1.out_conv.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet1.resConfUnit1.conv1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet1.resConfUnit1.conv1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet1.resConfUnit1.conv2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet1.resConfUnit1.conv2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet1.resConfUnit2.conv1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet1.resConfUnit2.conv1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet1.resConfUnit2.conv2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet1.resConfUnit2.conv2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet2.out_conv.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet2.out_conv.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet2.resConfUnit1.conv1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet2.resConfUnit1.conv1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet2.resConfUnit1.conv2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet2.resConfUnit1.conv2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet2.resConfUnit2.conv1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet2.resConfUnit2.conv1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet2.resConfUnit2.conv2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet2.resConfUnit2.conv2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet3.out_conv.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet3.out_conv.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet3.resConfUnit1.conv1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet3.resConfUnit1.conv1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet3.resConfUnit1.conv2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet3.resConfUnit1.conv2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet3.resConfUnit2.conv1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet3.resConfUnit2.conv1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet3.resConfUnit2.conv2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet3.resConfUnit2.conv2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet4.out_conv.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet4.out_conv.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet4.resConfUnit1.conv1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet4.resConfUnit1.conv1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet4.resConfUnit1.conv2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet4.resConfUnit1.conv2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet4.resConfUnit2.conv1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet4.resConfUnit2.conv1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet4.resConfUnit2.conv2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.refinenet4.resConfUnit2.conv2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.output_conv1.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.output_conv1.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.output_conv2.0.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.output_conv2.0.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.output_conv2.2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.core.core.depth_head.scratch.output_conv2.2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.conv2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.conv2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.seed_bin_regressor._net.0.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.seed_bin_regressor._net.0.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.seed_bin_regressor._net.2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.seed_bin_regressor._net.2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.seed_projector._net.0.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.seed_projector._net.0.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.seed_projector._net.2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.seed_projector._net.2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.projectors.0._net.0.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.projectors.0._net.0.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.projectors.0._net.2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.projectors.0._net.2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.projectors.1._net.0.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.projectors.1._net.0.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.projectors.1._net.2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.projectors.1._net.2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.projectors.2._net.0.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.projectors.2._net.0.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.projectors.2._net.2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.projectors.2._net.2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.projectors.3._net.0.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.projectors.3._net.0.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.projectors.3._net.2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.projectors.3._net.2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.attractors.0._net.0.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.attractors.0._net.0.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.attractors.0._net.2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.attractors.0._net.2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.attractors.1._net.0.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.attractors.1._net.0.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.attractors.1._net.2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.attractors.1._net.2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.attractors.2._net.0.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.attractors.2._net.0.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.attractors.2._net.2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.attractors.2._net.2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.attractors.3._net.0.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.attractors.3._net.0.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.attractors.3._net.2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.attractors.3._net.2.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.conditional_log_binomial.mlp.0.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.conditional_log_binomial.mlp.0.bias
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.conditional_log_binomial.mlp.2.weight
+2024/03/15 00:20:47 - patchstitcher - INFO - training param: module.coarse_branch.conditional_log_binomial.mlp.2.bias
+2024/03/15 00:23:05 - patchstitcher - INFO - Epoch: [01/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 2.218886375427246 - coarse_loss: 2.218886375427246
+2024/03/15 00:24:52 - patchstitcher - INFO - Epoch: [01/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 2.0132031440734863 - coarse_loss: 2.0132031440734863
+2024/03/15 00:26:41 - patchstitcher - INFO - Epoch: [01/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 2.1340489387512207 - coarse_loss: 2.1340489387512207
+2024/03/15 00:28:31 - patchstitcher - INFO - Epoch: [01/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.68356192111969 - coarse_loss: 1.68356192111969
+2024/03/15 00:31:46 - patchstitcher - INFO - Epoch: [02/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.1240144968032837 - coarse_loss: 1.1240144968032837
+2024/03/15 00:33:37 - patchstitcher - INFO - Epoch: [02/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.2552540302276611 - coarse_loss: 1.2552540302276611
+2024/03/15 00:35:27 - patchstitcher - INFO - Epoch: [02/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.3931670188903809 - coarse_loss: 1.3931670188903809
+2024/03/15 00:37:17 - patchstitcher - INFO - Epoch: [02/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.4315416812896729 - coarse_loss: 1.4315416812896729
+2024/03/15 00:38:56 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+----------+-----------+-----------+------------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |   rmse   |   log_10  |  rmse_log |   silog    |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+----------+-----------+-----------+------------+-----------+-----------+
+| 0.9222131 | 0.9841732 | 0.9937032 | 0.0942684 | 1.901311 | 0.0392215 | 0.1319014 | 11.5870857 | 0.3169146 | 1.4523976 |
++-----------+-----------+-----------+-----------+----------+-----------+-----------+------------+-----------+-----------+
+2024/03/15 00:40:53 - patchstitcher - INFO - Epoch: [03/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.3891286849975586 - coarse_loss: 1.3891286849975586
+2024/03/15 00:42:45 - patchstitcher - INFO - Epoch: [03/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.3853542804718018 - coarse_loss: 1.3853542804718018
+2024/03/15 00:44:31 - patchstitcher - INFO - Epoch: [03/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.6085820198059082 - coarse_loss: 1.6085820198059082
+2024/03/15 00:46:24 - patchstitcher - INFO - Epoch: [03/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.2743269205093384 - coarse_loss: 1.2743269205093384
+2024/03/15 00:49:33 - patchstitcher - INFO - Epoch: [04/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.4644969701766968 - coarse_loss: 1.4644969701766968
+2024/03/15 00:51:20 - patchstitcher - INFO - Epoch: [04/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.040415644645691 - coarse_loss: 1.040415644645691
+2024/03/15 00:53:07 - patchstitcher - INFO - Epoch: [04/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.2523736953735352 - coarse_loss: 1.2523736953735352
+2024/03/15 00:54:57 - patchstitcher - INFO - Epoch: [04/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.7893640995025635 - coarse_loss: 0.7893640995025635
+2024/03/15 00:56:31 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+----------+-----------+-----------+------------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |   rmse   |   log_10  |  rmse_log |   silog    |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+----------+-----------+-----------+------------+-----------+-----------+
+| 0.9466366 | 0.9857079 | 0.9944696 | 0.0784504 | 1.723246 | 0.0331783 | 0.1166779 | 10.4672395 | 0.2658952 | 1.2480133 |
++-----------+-----------+-----------+-----------+----------+-----------+-----------+------------+-----------+-----------+
+2024/03/15 00:58:25 - patchstitcher - INFO - Epoch: [05/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.8934182524681091 - coarse_loss: 0.8934182524681091
+2024/03/15 01:00:10 - patchstitcher - INFO - Epoch: [05/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.0365135669708252 - coarse_loss: 1.0365135669708252
+2024/03/15 01:02:00 - patchstitcher - INFO - Epoch: [05/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.0158889293670654 - coarse_loss: 1.0158889293670654
+2024/03/15 01:03:50 - patchstitcher - INFO - Epoch: [05/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.7366129159927368 - coarse_loss: 0.7366129159927368
+2024/03/15 01:07:04 - patchstitcher - INFO - Epoch: [06/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.4556183815002441 - coarse_loss: 1.4556183815002441
+2024/03/15 01:08:51 - patchstitcher - INFO - Epoch: [06/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.093213677406311 - coarse_loss: 1.093213677406311
+2024/03/15 01:10:45 - patchstitcher - INFO - Epoch: [06/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.8329901099205017 - coarse_loss: 0.8329901099205017
+2024/03/15 01:12:32 - patchstitcher - INFO - Epoch: [06/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.8255199193954468 - coarse_loss: 0.8255199193954468
+2024/03/15 01:14:05 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  |  rmse_log |   silog   |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+| 0.9492006 | 0.9876058 | 0.9947174 | 0.0765434 | 1.6623389 | 0.0336977 | 0.1157899 | 10.168448 | 0.2274059 | 1.1601292 |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+2024/03/15 01:16:01 - patchstitcher - INFO - Epoch: [07/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.9320656061172485 - coarse_loss: 0.9320656061172485
+2024/03/15 01:17:44 - patchstitcher - INFO - Epoch: [07/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.3558683395385742 - coarse_loss: 1.3558683395385742
+2024/03/15 01:19:36 - patchstitcher - INFO - Epoch: [07/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.1851251125335693 - coarse_loss: 1.1851251125335693
+2024/03/15 01:21:28 - patchstitcher - INFO - Epoch: [07/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.7694360613822937 - coarse_loss: 0.7694360613822937
+2024/03/15 01:24:39 - patchstitcher - INFO - Epoch: [08/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.9268642067909241 - coarse_loss: 0.9268642067909241
+2024/03/15 01:26:28 - patchstitcher - INFO - Epoch: [08/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.0070387125015259 - coarse_loss: 1.0070387125015259
+2024/03/15 01:28:17 - patchstitcher - INFO - Epoch: [08/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.3363308906555176 - coarse_loss: 1.3363308906555176
+2024/03/15 01:30:03 - patchstitcher - INFO - Epoch: [08/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.549015998840332 - coarse_loss: 1.549015998840332
+2024/03/15 01:31:38 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  |  rmse_log |   silog   |  sq_rel  |    see    |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+----------+-----------+
+| 0.9580896 | 0.9882235 | 0.9949475 | 0.0697348 | 1.6023046 | 0.0295156 | 0.1067427 | 9.6755001 | 0.224005 | 1.1545794 |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+----------+-----------+
+2024/03/15 01:33:33 - patchstitcher - INFO - Epoch: [09/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.2725939750671387 - coarse_loss: 1.2725939750671387
+2024/03/15 01:35:25 - patchstitcher - INFO - Epoch: [09/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.8319188356399536 - coarse_loss: 1.8319188356399536
+2024/03/15 01:37:16 - patchstitcher - INFO - Epoch: [09/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.9146164655685425 - coarse_loss: 0.9146164655685425
+2024/03/15 01:39:08 - patchstitcher - INFO - Epoch: [09/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.9933633208274841 - coarse_loss: 0.9933633208274841
+2024/03/15 01:42:21 - patchstitcher - INFO - Epoch: [10/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.531670331954956 - coarse_loss: 0.531670331954956
+2024/03/15 01:44:13 - patchstitcher - INFO - Epoch: [10/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.98005211353302 - coarse_loss: 0.98005211353302
+2024/03/15 01:46:08 - patchstitcher - INFO - Epoch: [10/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.562972068786621 - coarse_loss: 1.562972068786621
+2024/03/15 01:48:00 - patchstitcher - INFO - Epoch: [10/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.0578055381774902 - coarse_loss: 1.0578055381774902
+2024/03/15 01:49:39 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+----------+-----------+--------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |  log_10  |  rmse_log | silog  |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+-----------+----------+-----------+--------+-----------+-----------+
+| 0.9573399 | 0.9884624 | 0.9949526 | 0.0727779 | 1.5619678 | 0.030998 | 0.1089102 | 9.5524 | 0.2075647 | 1.1259904 |
++-----------+-----------+-----------+-----------+-----------+----------+-----------+--------+-----------+-----------+
+2024/03/15 01:51:35 - patchstitcher - INFO - Epoch: [11/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.9536416530609131 - coarse_loss: 0.9536416530609131
+2024/03/15 01:53:33 - patchstitcher - INFO - Epoch: [11/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.061272382736206 - coarse_loss: 1.061272382736206
+2024/03/15 01:55:27 - patchstitcher - INFO - Epoch: [11/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.403846263885498 - coarse_loss: 1.403846263885498
+2024/03/15 01:57:19 - patchstitcher - INFO - Epoch: [11/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.6634379625320435 - coarse_loss: 0.6634379625320435
+2024/03/15 02:00:39 - patchstitcher - INFO - Epoch: [12/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.7105982303619385 - coarse_loss: 0.7105982303619385
+2024/03/15 02:02:34 - patchstitcher - INFO - Epoch: [12/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.8706010580062866 - coarse_loss: 0.8706010580062866
+2024/03/15 02:04:29 - patchstitcher - INFO - Epoch: [12/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.013525366783142 - coarse_loss: 1.013525366783142
+2024/03/15 02:06:17 - patchstitcher - INFO - Epoch: [12/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.9357657432556152 - coarse_loss: 0.9357657432556152
+2024/03/15 02:07:50 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  |  rmse_log |   silog   |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+| 0.9572283 | 0.9886936 | 0.9950871 | 0.0731142 | 1.5668887 | 0.0308406 | 0.1077591 | 9.5263897 | 0.2176418 | 1.1755943 |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+2024/03/15 02:09:47 - patchstitcher - INFO - Epoch: [13/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.4000838994979858 - coarse_loss: 1.4000838994979858
+2024/03/15 02:11:41 - patchstitcher - INFO - Epoch: [13/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.147301435470581 - coarse_loss: 1.147301435470581
+2024/03/15 02:13:39 - patchstitcher - INFO - Epoch: [13/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.9417784214019775 - coarse_loss: 0.9417784214019775
+2024/03/15 02:15:36 - patchstitcher - INFO - Epoch: [13/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.898872971534729 - coarse_loss: 0.898872971534729
+2024/03/15 02:18:53 - patchstitcher - INFO - Epoch: [14/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.6218137741088867 - coarse_loss: 0.6218137741088867
+2024/03/15 02:20:49 - patchstitcher - INFO - Epoch: [14/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.9591147303581238 - coarse_loss: 0.9591147303581238
+2024/03/15 02:22:42 - patchstitcher - INFO - Epoch: [14/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.7330798506736755 - coarse_loss: 0.7330798506736755
+2024/03/15 02:24:37 - patchstitcher - INFO - Epoch: [14/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.671249508857727 - coarse_loss: 0.671249508857727
+2024/03/15 02:26:12 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  |  rmse_log |   silog   |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+| 0.9656246 | 0.9891472 | 0.9951051 | 0.0614303 | 1.5103214 | 0.0264747 | 0.1001097 | 9.4011921 | 0.1946697 | 1.0991172 |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+2024/03/15 02:28:10 - patchstitcher - INFO - Epoch: [15/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.7798411846160889 - coarse_loss: 0.7798411846160889
+2024/03/15 02:30:02 - patchstitcher - INFO - Epoch: [15/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.9757544994354248 - coarse_loss: 0.9757544994354248
+2024/03/15 02:31:49 - patchstitcher - INFO - Epoch: [15/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.1485944986343384 - coarse_loss: 1.1485944986343384
+2024/03/15 02:33:42 - patchstitcher - INFO - Epoch: [15/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.8730670809745789 - coarse_loss: 0.8730670809745789
+2024/03/15 02:36:57 - patchstitcher - INFO - Epoch: [16/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.0859500169754028 - coarse_loss: 1.0859500169754028
+2024/03/15 02:38:46 - patchstitcher - INFO - Epoch: [16/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.9123729467391968 - coarse_loss: 0.9123729467391968
+2024/03/15 02:40:36 - patchstitcher - INFO - Epoch: [16/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.0700657367706299 - coarse_loss: 1.0700657367706299
+2024/03/15 02:42:24 - patchstitcher - INFO - Epoch: [16/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.8393980264663696 - coarse_loss: 1.8393980264663696
+2024/03/15 02:43:58 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+----------+-----------+-----------+-----------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |   rmse   |   log_10  |  rmse_log |   silog   |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+----------+-----------+-----------+-----------+-----------+-----------+
+| 0.9678091 | 0.9892931 | 0.9952321 | 0.0607629 | 1.488932 | 0.0257705 | 0.0981663 | 9.0934609 | 0.1966839 | 1.0878515 |
++-----------+-----------+-----------+-----------+----------+-----------+-----------+-----------+-----------+-----------+
+2024/03/15 02:45:51 - patchstitcher - INFO - Epoch: [17/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.7053128480911255 - coarse_loss: 0.7053128480911255
+2024/03/15 02:47:43 - patchstitcher - INFO - Epoch: [17/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.9886703491210938 - coarse_loss: 0.9886703491210938
+2024/03/15 02:49:32 - patchstitcher - INFO - Epoch: [17/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.180053949356079 - coarse_loss: 1.180053949356079
+2024/03/15 02:51:22 - patchstitcher - INFO - Epoch: [17/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.316230297088623 - coarse_loss: 1.316230297088623
+2024/03/15 02:54:39 - patchstitcher - INFO - Epoch: [18/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.7665231227874756 - coarse_loss: 0.7665231227874756
+2024/03/15 02:56:30 - patchstitcher - INFO - Epoch: [18/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.6590834856033325 - coarse_loss: 0.6590834856033325
+2024/03/15 02:58:17 - patchstitcher - INFO - Epoch: [18/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.9268083572387695 - coarse_loss: 0.9268083572387695
+2024/03/15 03:00:07 - patchstitcher - INFO - Epoch: [18/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.255874752998352 - coarse_loss: 1.255874752998352
+2024/03/15 03:01:42 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  |  rmse_log |   silog   |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+| 0.9691702 | 0.9894969 | 0.9952754 | 0.0559551 | 1.4743834 | 0.0240017 | 0.0943829 | 8.8561864 | 0.1819411 | 1.0395958 |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+2024/03/15 03:03:35 - patchstitcher - INFO - Epoch: [19/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.6205350756645203 - coarse_loss: 0.6205350756645203
+2024/03/15 03:05:29 - patchstitcher - INFO - Epoch: [19/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.6529569625854492 - coarse_loss: 0.6529569625854492
+2024/03/15 03:07:18 - patchstitcher - INFO - Epoch: [19/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.8907508850097656 - coarse_loss: 0.8907508850097656
+2024/03/15 03:09:13 - patchstitcher - INFO - Epoch: [19/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.5823774337768555 - coarse_loss: 0.5823774337768555
+2024/03/15 03:12:22 - patchstitcher - INFO - Epoch: [20/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.3379265069961548 - coarse_loss: 1.3379265069961548
+2024/03/15 03:14:13 - patchstitcher - INFO - Epoch: [20/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.615516185760498 - coarse_loss: 0.615516185760498
+2024/03/15 03:16:04 - patchstitcher - INFO - Epoch: [20/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.5864847302436829 - coarse_loss: 0.5864847302436829
+2024/03/15 03:17:54 - patchstitcher - INFO - Epoch: [20/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.9669459462165833 - coarse_loss: 0.9669459462165833
+2024/03/15 03:19:29 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  |  rmse_log |   silog   |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+| 0.9697102 | 0.9895742 | 0.9953449 | 0.0539071 | 1.4497501 | 0.0229091 | 0.0925752 | 8.7784555 | 0.1802817 | 1.0580258 |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+2024/03/15 03:21:25 - patchstitcher - INFO - Epoch: [21/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.9557666778564453 - coarse_loss: 0.9557666778564453
+2024/03/15 03:23:15 - patchstitcher - INFO - Epoch: [21/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.6958411931991577 - coarse_loss: 0.6958411931991577
+2024/03/15 03:25:01 - patchstitcher - INFO - Epoch: [21/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.5607629418373108 - coarse_loss: 0.5607629418373108
+2024/03/15 03:26:54 - patchstitcher - INFO - Epoch: [21/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.8118071556091309 - coarse_loss: 1.8118071556091309
+2024/03/15 03:30:05 - patchstitcher - INFO - Epoch: [22/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.0183720588684082 - coarse_loss: 1.0183720588684082
+2024/03/15 03:31:53 - patchstitcher - INFO - Epoch: [22/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.0083253383636475 - coarse_loss: 1.0083253383636475
+2024/03/15 03:33:45 - patchstitcher - INFO - Epoch: [22/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.5852430462837219 - coarse_loss: 0.5852430462837219
+2024/03/15 03:35:35 - patchstitcher - INFO - Epoch: [22/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.8135958909988403 - coarse_loss: 0.8135958909988403
+2024/03/15 03:37:10 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  |  rmse_log |   silog   |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+| 0.9699838 | 0.9896694 | 0.9953757 | 0.0526183 | 1.4463599 | 0.0224501 | 0.0915034 | 8.7251583 | 0.1771403 | 1.0479052 |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+
+2024/03/15 03:39:04 - patchstitcher - INFO - Epoch: [23/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.6198912262916565 - coarse_loss: 0.6198912262916565
+2024/03/15 03:40:57 - patchstitcher - INFO - Epoch: [23/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.1995759010314941 - coarse_loss: 1.1995759010314941
+2024/03/15 03:42:47 - patchstitcher - INFO - Epoch: [23/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.7696393728256226 - coarse_loss: 1.7696393728256226
+2024/03/15 03:44:34 - patchstitcher - INFO - Epoch: [23/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.4660639762878418 - coarse_loss: 1.4660639762878418
+2024/03/15 03:47:48 - patchstitcher - INFO - Epoch: [24/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.9167467355728149 - coarse_loss: 0.9167467355728149
+2024/03/15 03:49:40 - patchstitcher - INFO - Epoch: [24/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.6638955473899841 - coarse_loss: 0.6638955473899841
+2024/03/15 03:51:31 - patchstitcher - INFO - Epoch: [24/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.4969237446784973 - coarse_loss: 0.4969237446784973
+2024/03/15 03:53:18 - patchstitcher - INFO - Epoch: [24/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.5059656500816345 - coarse_loss: 0.5059656500816345
+2024/03/15 03:54:52 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  |  rmse_log |   silog   |   sq_rel  |   see    |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+----------+
+| 0.9701259 | 0.9896677 | 0.9953767 | 0.0521426 | 1.4457442 | 0.0222779 | 0.0914182 | 8.7319618 | 0.1776046 | 1.046505 |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+----------+
+2024/03/15 03:54:52 - patchstitcher - INFO - Saving ckp, but use the inner get_save_dict fuction to get model_dict
+2024/03/15 03:54:52 - patchstitcher - INFO - For saving space. Would you like to save base model several times? :>
+2024/03/15 03:54:52 - patchstitcher - INFO - save checkpoint_24.pth at ./work_dir/depthanything_vits_u4k/coarse_pretrain

depthanything_vits_u4k/coarse_pretrain/checkpoint_24.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b1eaf830589ea843ca421c175967b31c33609616f04af50b18b40d1abb3cb1e3
+size 300162730

depthanything_vits_u4k/coarse_pretrain/config.py

@@ -0,0 +1,310 @@
+collect_input_args = [
+    'image_lr',
+    'crops_image_hr',
+    'depth_gt',
+    'crop_depths',
+    'bboxs',
+    'image_hr',
+]
+convert_syncbn = True
+debug = False
+env_cfg = dict(
+    cudnn_benchmark=True,
+    dist_cfg=dict(backend='nccl'),
+    mp_cfg=dict(mp_start_method='forkserver'))
+find_unused_parameters = True
+general_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        dataset_name='', gt_dir=None, rgb_image_dir='', type='ImageDataset'),
+    num_workers=2)
+launcher = 'pytorch'
+log_name = 'coarse_pretrain'
+max_depth = 80
+min_depth = 0.001
+model = dict(
+    coarse_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        depth_anything=True,
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            392,
+            518,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='vits',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource='local::./work_dir/DepthAnything_vits.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='DA-ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    fine_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        depth_anything=True,
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            392,
+            518,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='vits',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource='local::./work_dir/DepthAnything_vits.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='DA-ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    max_depth=80,
+    min_depth=0.001,
+    sigloss=dict(type='SILogLoss'),
+    target='coarse',
+    type='BaselinePretrain')
+optim_wrapper = dict(
+    clip_grad=dict(max_norm=0.1, norm_type=2, type='norm'),
+    optimizer=dict(lr=4e-06, type='AdamW', weight_decay=0.01),
+    paramwise_cfg=dict(bypass_duplicate=True, custom_keys=dict()))
+param_scheduler = dict(
+    base_momentum=0.85,
+    cycle_momentum=True,
+    div_factor=1,
+    final_div_factor=10000,
+    max_momentum=0.95,
+    pct_start=0.5,
+    three_phase=False)
+project = 'patchfusion'
+resume = False
+tags = [
+    'coarse',
+    'da',
+    'vits',
+]
+test_in_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+test_out_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test_out.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+train_cfg = dict(
+    eval_start=0,
+    log_interval=100,
+    max_epochs=24,
+    save_checkpoint_interval=24,
+    train_log_img_interval=100,
+    val_interval=2,
+    val_log_img_interval=50,
+    val_type='epoch_base')
+train_dataloader = dict(
+    batch_size=4,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='train',
+        resize_mode='depth-anything',
+        split='./data/u4k/splits/train.txt',
+        transform_cfg=dict(
+            degree=1.0,
+            network_process_size=[
+                392,
+                518,
+            ],
+            random_crop=True,
+            random_crop_size=(
+                540,
+                960,
+            )),
+        type='UnrealStereo4kDataset'),
+    num_workers=4)
+val_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        resize_mode='depth-anything',
+        split='./data/u4k/splits/val.txt',
+        transform_cfg=dict(
+            degree=1.0,
+            network_process_size=[
+                392,
+                518,
+            ],
+            random_crop_size=(
+                540,
+                960,
+            )),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+work_dir = './work_dir/depthanything_vits_u4k/coarse_pretrain'
+zoe_depth_config = dict(
+    attractor_alpha=1000,
+    attractor_gamma=2,
+    attractor_kind='mean',
+    attractor_type='inv',
+    aug=True,
+    bin_centers_type='softplus',
+    bin_embedding_dim=128,
+    clip_grad=0.1,
+    dataset='nyu',
+    depth_anything=True,
+    distributed=True,
+    do_resize=False,
+    force_keep_ar=True,
+    freeze_midas_bn=True,
+    gpu='NULL',
+    img_size=[
+        392,
+        518,
+    ],
+    inverse_midas=False,
+    log_images_every=0.1,
+    max_depth=80,
+    max_temp=50.0,
+    max_translation=100,
+    memory_efficient=True,
+    midas_model_type='vits',
+    min_depth=0.001,
+    min_temp=0.0212,
+    model='zoedepth',
+    n_attractors=[
+        16,
+        8,
+        4,
+        1,
+    ],
+    n_bins=64,
+    name='ZoeDepth',
+    notes='',
+    output_distribution='logbinomial',
+    prefetch=False,
+    pretrained_resource='local::./work_dir/DepthAnything_vits.pt',
+    print_losses=False,
+    project='ZoeDepth',
+    random_crop=False,
+    random_translate=False,
+    root='.',
+    save_dir='',
+    shared_dict='NULL',
+    tags='',
+    train_midas=True,
+    translate_prob=0.2,
+    type='DA-ZoeDepth',
+    uid='NULL',
+    use_amp=False,
+    use_pretrained_midas=True,
+    use_shared_dict=False,
+    validate_every=0.25,
+    version_name='v1',
+    workers=16)

depthanything_vits_u4k/fine_pretrain/20240315_035516.log

@@ -0,0 +1,1028 @@
+2024/03/15 03:55:26 - patchstitcher - INFO - 
+------------------------------------------------------------
+System environment:
+    sys.platform: linux
+    Python: 3.8.18 | packaged by conda-forge | (default, Oct 10 2023, 15:44:36) [GCC 12.3.0]
+    CUDA available: True
+    numpy_random_seed: 621
+    GPU 0,1,2,3: NVIDIA A100-SXM4-80GB
+    CUDA_HOME: /sw/rl9g/cuda/11.8/rl9_binary
+    NVCC: Cuda compilation tools, release 11.8, V11.8.89
+    GCC: gcc (GCC) 11.3.1 20220421 (Red Hat 11.3.1-2)
+    PyTorch: 2.1.2
+    PyTorch compiling details: PyTorch built with:
+  - GCC 9.3
+  - C++ Version: 201703
+  - Intel(R) oneAPI Math Kernel Library Version 2022.1-Product Build 20220311 for Intel(R) 64 architecture applications
+  - Intel(R) MKL-DNN v3.1.1 (Git Hash 64f6bcbcbab628e96f33a62c3e975f8535a7bde4)
+  - OpenMP 201511 (a.k.a. OpenMP 4.5)
+  - LAPACK is enabled (usually provided by MKL)
+  - NNPACK is enabled
+  - CPU capability usage: AVX2
+  - CUDA Runtime 11.8
+  - NVCC architecture flags: -gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_61,code=sm_61;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_80,code=sm_80;-gencode;arch=compute_86,code=sm_86;-gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_90,code=sm_90;-gencode;arch=compute_37,code=compute_37
+  - CuDNN 8.7
+  - Magma 2.6.1
+  - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=11.8, CUDNN_VERSION=8.7.0, CXX_COMPILER=/opt/rh/devtoolset-9/root/usr/bin/c++, CXX_FLAGS= -D_GLIBCXX_USE_CXX11_ABI=0 -fabi-version=11 -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOROCTRACER -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -O2 -fPIC -Wall -Wextra -Werror=return-type -Werror=non-virtual-dtor -Werror=bool-operation -Wnarrowing -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wno-unused-parameter -Wno-unused-function -Wno-unused-result -Wno-strict-overflow -Wno-strict-aliasing -Wno-stringop-overflow -Wno-psabi -Wno-error=pedantic -Wno-error=old-style-cast -Wno-invalid-partial-specialization -Wno-unused-private-field -Wno-aligned-allocation-unavailable -Wno-missing-braces -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Werror=cast-function-type -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_DISABLE_GPU_ASSERTS=ON, TORCH_VERSION=2.1.2, USE_CUDA=ON, USE_CUDNN=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=ON, USE_NNPACK=ON, USE_OPENMP=ON, USE_ROCM=OFF, 
+
+    TorchVision: 0.16.2
+    OpenCV: 4.8.1
+    MMEngine: 0.10.2
+
+Runtime environment:
+    cudnn_benchmark: True
+    mp_cfg: {'mp_start_method': 'forkserver'}
+    dist_cfg: {'backend': 'nccl'}
+    seed: 621
+    Distributed launcher: pytorch
+    Distributed training: True
+    GPU number: 4
+------------------------------------------------------------
+
+2024/03/15 03:55:26 - patchstitcher - INFO - Config:
+collect_input_args = [
+    'image_lr',
+    'crops_image_hr',
+    'depth_gt',
+    'crop_depths',
+    'bboxs',
+    'image_hr',
+]
+convert_syncbn = True
+debug = False
+env_cfg = dict(
+    cudnn_benchmark=True,
+    dist_cfg=dict(backend='nccl'),
+    mp_cfg=dict(mp_start_method='forkserver'))
+find_unused_parameters = True
+general_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        dataset_name='', gt_dir=None, rgb_image_dir='', type='ImageDataset'),
+    num_workers=2)
+launcher = 'pytorch'
+log_name = 'fine_pretrain'
+max_depth = 80
+min_depth = 0.001
+model = dict(
+    coarse_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        depth_anything=True,
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            392,
+            518,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='vits',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource='local::./work_dir/DepthAnything_vits.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='DA-ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    fine_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        depth_anything=True,
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            392,
+            518,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='vits',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource='local::./work_dir/DepthAnything_vits.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='DA-ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    max_depth=80,
+    min_depth=0.001,
+    patch_process_shape=(
+        392,
+        518,
+    ),
+    sigloss=dict(type='SILogLoss'),
+    target='fine',
+    type='BaselinePretrain')
+optim_wrapper = dict(
+    clip_grad=dict(max_norm=0.1, norm_type=2, type='norm'),
+    optimizer=dict(lr=4e-06, type='AdamW', weight_decay=0.01),
+    paramwise_cfg=dict(bypass_duplicate=True, custom_keys=dict()))
+param_scheduler = dict(
+    base_momentum=0.85,
+    cycle_momentum=True,
+    div_factor=1,
+    final_div_factor=10000,
+    max_momentum=0.95,
+    pct_start=0.5,
+    three_phase=False)
+project = 'patchfusion'
+tags = [
+    'fine',
+    'da',
+    'vits',
+]
+test_in_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+test_out_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test_out.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+train_cfg = dict(
+    eval_start=0,
+    log_interval=100,
+    max_epochs=24,
+    save_checkpoint_interval=24,
+    train_log_img_interval=100,
+    val_interval=2,
+    val_log_img_interval=50,
+    val_type='epoch_base')
+train_dataloader = dict(
+    batch_size=4,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='train',
+        resize_mode='depth-anything',
+        split='./data/u4k/splits/train.txt',
+        transform_cfg=dict(
+            degree=1.0, network_process_size=[
+                392,
+                518,
+            ], random_crop=True),
+        type='UnrealStereo4kDataset'),
+    num_workers=4)
+val_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        resize_mode='depth-anything',
+        split='./data/u4k/splits/val.txt',
+        transform_cfg=dict(degree=1.0, network_process_size=[
+            392,
+            518,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+work_dir = './work_dir/depthanything_vits_u4k/fine_pretrain'
+zoe_depth_config = dict(
+    attractor_alpha=1000,
+    attractor_gamma=2,
+    attractor_kind='mean',
+    attractor_type='inv',
+    aug=True,
+    bin_centers_type='softplus',
+    bin_embedding_dim=128,
+    clip_grad=0.1,
+    dataset='nyu',
+    depth_anything=True,
+    distributed=True,
+    do_resize=False,
+    force_keep_ar=True,
+    freeze_midas_bn=True,
+    gpu='NULL',
+    img_size=[
+        392,
+        518,
+    ],
+    inverse_midas=False,
+    log_images_every=0.1,
+    max_depth=80,
+    max_temp=50.0,
+    max_translation=100,
+    memory_efficient=True,
+    midas_model_type='vits',
+    min_depth=0.001,
+    min_temp=0.0212,
+    model='zoedepth',
+    n_attractors=[
+        16,
+        8,
+        4,
+        1,
+    ],
+    n_bins=64,
+    name='ZoeDepth',
+    notes='',
+    output_distribution='logbinomial',
+    prefetch=False,
+    pretrained_resource='local::./work_dir/DepthAnything_vits.pt',
+    print_losses=False,
+    project='ZoeDepth',
+    random_crop=False,
+    random_translate=False,
+    root='.',
+    save_dir='',
+    shared_dict='NULL',
+    tags='',
+    train_midas=True,
+    translate_prob=0.2,
+    type='DA-ZoeDepth',
+    uid='NULL',
+    use_amp=False,
+    use_pretrained_midas=True,
+    use_shared_dict=False,
+    validate_every=0.25,
+    version_name='v1',
+    workers=16)
+
+2024/03/15 03:55:27 - patchstitcher - INFO - Loading deepnet from local::./work_dir/DepthAnything_vits.pt
+2024/03/15 03:55:27 - patchstitcher - INFO - Current zoedepth.core.prep.resizer is <class 'torch.nn.modules.linear.Identity'>
+2024/03/15 03:55:27 - patchstitcher - INFO - DistributedDataParallel(
+  (module): BaselinePretrain(
+    (fine_branch): ZoeDepth(
+      (core): DepthAnythingCore(
+        (core): DPT_DINOv2(
+          (pretrained): DinoVisionTransformer(
+            (patch_embed): PatchEmbed(
+              (proj): Conv2d(3, 384, kernel_size=(14, 14), stride=(14, 14))
+              (norm): Identity()
+            )
+            (blocks): ModuleList(
+              (0-11): 12 x NestedTensorBlock(
+                (norm1): LayerNorm((384,), eps=1e-06, elementwise_affine=True)
+                (attn): MemEffAttention(
+                  (qkv): Linear(in_features=384, out_features=1152, bias=True)
+                  (attn_drop): Dropout(p=0.0, inplace=False)
+                  (proj): Linear(in_features=384, out_features=384, bias=True)
+                  (proj_drop): Dropout(p=0.0, inplace=False)
+                )
+                (ls1): LayerScale()
+                (drop_path1): Identity()
+                (norm2): LayerNorm((384,), eps=1e-06, elementwise_affine=True)
+                (mlp): Mlp(
+                  (fc1): Linear(in_features=384, out_features=1536, bias=True)
+                  (act): GELU(approximate='none')
+                  (fc2): Linear(in_features=1536, out_features=384, bias=True)
+                  (drop): Dropout(p=0.0, inplace=False)
+                )
+                (ls2): LayerScale()
+                (drop_path2): Identity()
+              )
+            )
+            (norm): LayerNorm((384,), eps=1e-06, elementwise_affine=True)
+            (head): Identity()
+          )
+          (depth_head): DPTHead(
+            (projects): ModuleList(
+              (0): Conv2d(384, 48, kernel_size=(1, 1), stride=(1, 1))
+              (1): Conv2d(384, 96, kernel_size=(1, 1), stride=(1, 1))
+              (2): Conv2d(384, 192, kernel_size=(1, 1), stride=(1, 1))
+              (3): Conv2d(384, 384, kernel_size=(1, 1), stride=(1, 1))
+            )
+            (resize_layers): ModuleList(
+              (0): ConvTranspose2d(48, 48, kernel_size=(4, 4), stride=(4, 4))
+              (1): ConvTranspose2d(96, 96, kernel_size=(2, 2), stride=(2, 2))
+              (2): Identity()
+              (3): Conv2d(384, 384, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1))
+            )
+            (scratch): Module(
+              (layer1_rn): Conv2d(48, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
+              (layer2_rn): Conv2d(96, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
+              (layer3_rn): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
+              (layer4_rn): Conv2d(384, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
+              (refinenet1): FeatureFusionBlock(
+                (out_conv): Conv2d(64, 64, kernel_size=(1, 1), stride=(1, 1))
+                (resConfUnit1): ResidualConvUnit(
+                  (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (resConfUnit2): ResidualConvUnit(
+                  (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (skip_add): FloatFunctional(
+                  (activation_post_process): Identity()
+                )
+              )
+              (refinenet2): FeatureFusionBlock(
+                (out_conv): Conv2d(64, 64, kernel_size=(1, 1), stride=(1, 1))
+                (resConfUnit1): ResidualConvUnit(
+                  (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (resConfUnit2): ResidualConvUnit(
+                  (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (skip_add): FloatFunctional(
+                  (activation_post_process): Identity()
+                )
+              )
+              (refinenet3): FeatureFusionBlock(
+                (out_conv): Conv2d(64, 64, kernel_size=(1, 1), stride=(1, 1))
+                (resConfUnit1): ResidualConvUnit(
+                  (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (resConfUnit2): ResidualConvUnit(
+                  (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (skip_add): FloatFunctional(
+                  (activation_post_process): Identity()
+                )
+              )
+              (refinenet4): FeatureFusionBlock(
+                (out_conv): Conv2d(64, 64, kernel_size=(1, 1), stride=(1, 1))
+                (resConfUnit1): ResidualConvUnit(
+                  (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (resConfUnit2): ResidualConvUnit(
+                  (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                  (activation): ReLU()
+                  (skip_add): FloatFunctional(
+                    (activation_post_process): Identity()
+                  )
+                )
+                (skip_add): FloatFunctional(
+                  (activation_post_process): Identity()
+                )
+              )
+              (output_conv1): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+              (output_conv2): Sequential(
+                (0): Conv2d(32, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
+                (1): ReLU(inplace=True)
+                (2): Conv2d(32, 1, kernel_size=(1, 1), stride=(1, 1))
+                (3): ReLU(inplace=True)
+                (4): Identity()
+              )
+            )
+          )
+        )
+      )
+      (conv2): Conv2d(64, 64, kernel_size=(1, 1), stride=(1, 1))
+      (seed_bin_regressor): SeedBinRegressorUnnormed(
+        (_net): Sequential(
+          (0): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1))
+          (1): ReLU(inplace=True)
+          (2): Conv2d(256, 64, kernel_size=(1, 1), stride=(1, 1))
+          (3): Softplus(beta=1, threshold=20)
+        )
+      )
+      (seed_projector): Projector(
+        (_net): Sequential(
+          (0): Conv2d(64, 128, kernel_size=(1, 1), stride=(1, 1))
+          (1): ReLU(inplace=True)
+          (2): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+        )
+      )
+      (projectors): ModuleList(
+        (0-3): 4 x Projector(
+          (_net): Sequential(
+            (0): Conv2d(64, 128, kernel_size=(1, 1), stride=(1, 1))
+            (1): ReLU(inplace=True)
+            (2): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+          )
+        )
+      )
+      (attractors): ModuleList(
+        (0): AttractorLayerUnnormed(
+          (_net): Sequential(
+            (0): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+            (1): ReLU(inplace=True)
+            (2): Conv2d(128, 16, kernel_size=(1, 1), stride=(1, 1))
+            (3): Softplus(beta=1, threshold=20)
+          )
+        )
+        (1): AttractorLayerUnnormed(
+          (_net): Sequential(
+            (0): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+            (1): ReLU(inplace=True)
+            (2): Conv2d(128, 8, kernel_size=(1, 1), stride=(1, 1))
+            (3): Softplus(beta=1, threshold=20)
+          )
+        )
+        (2): AttractorLayerUnnormed(
+          (_net): Sequential(
+            (0): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+            (1): ReLU(inplace=True)
+            (2): Conv2d(128, 4, kernel_size=(1, 1), stride=(1, 1))
+            (3): Softplus(beta=1, threshold=20)
+          )
+        )
+        (3): AttractorLayerUnnormed(
+          (_net): Sequential(
+            (0): Conv2d(128, 128, kernel_size=(1, 1), stride=(1, 1))
+            (1): ReLU(inplace=True)
+            (2): Conv2d(128, 1, kernel_size=(1, 1), stride=(1, 1))
+            (3): Softplus(beta=1, threshold=20)
+          )
+        )
+      )
+      (conditional_log_binomial): ConditionalLogBinomial(
+        (log_binomial_transform): LogBinomial()
+        (mlp): Sequential(
+          (0): Conv2d(161, 80, kernel_size=(1, 1), stride=(1, 1))
+          (1): GELU(approximate='none')
+          (2): Conv2d(80, 4, kernel_size=(1, 1), stride=(1, 1))
+          (3): Softplus(beta=1, threshold=20)
+        )
+      )
+    )
+    (sigloss): SILogLoss()
+  )
+)
+2024/03/15 03:55:33 - patchstitcher - INFO - successfully init trainer
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.cls_token
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.pos_embed
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.mask_token
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.patch_embed.proj.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.patch_embed.proj.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.0.norm1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.0.norm1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.0.attn.qkv.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.0.attn.qkv.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.0.attn.proj.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.0.attn.proj.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.0.ls1.gamma
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.0.norm2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.0.norm2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.0.mlp.fc1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.0.mlp.fc1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.0.mlp.fc2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.0.mlp.fc2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.0.ls2.gamma
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.1.norm1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.1.norm1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.1.attn.qkv.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.1.attn.qkv.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.1.attn.proj.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.1.attn.proj.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.1.ls1.gamma
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.1.norm2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.1.norm2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.1.mlp.fc1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.1.mlp.fc1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.1.mlp.fc2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.1.mlp.fc2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.1.ls2.gamma
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.2.norm1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.2.norm1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.2.attn.qkv.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.2.attn.qkv.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.2.attn.proj.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.2.attn.proj.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.2.ls1.gamma
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.2.norm2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.2.norm2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.2.mlp.fc1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.2.mlp.fc1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.2.mlp.fc2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.2.mlp.fc2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.2.ls2.gamma
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.3.norm1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.3.norm1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.3.attn.qkv.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.3.attn.qkv.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.3.attn.proj.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.3.attn.proj.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.3.ls1.gamma
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.3.norm2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.3.norm2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.3.mlp.fc1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.3.mlp.fc1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.3.mlp.fc2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.3.mlp.fc2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.3.ls2.gamma
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.4.norm1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.4.norm1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.4.attn.qkv.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.4.attn.qkv.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.4.attn.proj.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.4.attn.proj.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.4.ls1.gamma
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.4.norm2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.4.norm2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.4.mlp.fc1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.4.mlp.fc1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.4.mlp.fc2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.4.mlp.fc2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.4.ls2.gamma
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.5.norm1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.5.norm1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.5.attn.qkv.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.5.attn.qkv.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.5.attn.proj.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.5.attn.proj.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.5.ls1.gamma
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.5.norm2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.5.norm2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.5.mlp.fc1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.5.mlp.fc1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.5.mlp.fc2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.5.mlp.fc2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.5.ls2.gamma
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.6.norm1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.6.norm1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.6.attn.qkv.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.6.attn.qkv.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.6.attn.proj.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.6.attn.proj.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.6.ls1.gamma
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.6.norm2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.6.norm2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.6.mlp.fc1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.6.mlp.fc1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.6.mlp.fc2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.6.mlp.fc2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.6.ls2.gamma
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.7.norm1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.7.norm1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.7.attn.qkv.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.7.attn.qkv.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.7.attn.proj.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.7.attn.proj.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.7.ls1.gamma
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.7.norm2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.7.norm2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.7.mlp.fc1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.7.mlp.fc1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.7.mlp.fc2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.7.mlp.fc2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.7.ls2.gamma
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.8.norm1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.8.norm1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.8.attn.qkv.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.8.attn.qkv.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.8.attn.proj.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.8.attn.proj.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.8.ls1.gamma
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.8.norm2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.8.norm2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.8.mlp.fc1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.8.mlp.fc1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.8.mlp.fc2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.8.mlp.fc2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.8.ls2.gamma
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.9.norm1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.9.norm1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.9.attn.qkv.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.9.attn.qkv.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.9.attn.proj.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.9.attn.proj.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.9.ls1.gamma
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.9.norm2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.9.norm2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.9.mlp.fc1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.9.mlp.fc1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.9.mlp.fc2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.9.mlp.fc2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.9.ls2.gamma
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.10.norm1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.10.norm1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.10.attn.qkv.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.10.attn.qkv.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.10.attn.proj.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.10.attn.proj.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.10.ls1.gamma
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.10.norm2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.10.norm2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.10.mlp.fc1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.10.mlp.fc1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.10.mlp.fc2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.10.mlp.fc2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.10.ls2.gamma
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.11.norm1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.11.norm1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.11.attn.qkv.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.11.attn.qkv.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.11.attn.proj.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.11.attn.proj.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.11.ls1.gamma
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.11.norm2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.11.norm2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.11.mlp.fc1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.11.mlp.fc1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.11.mlp.fc2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.11.mlp.fc2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.blocks.11.ls2.gamma
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.norm.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.pretrained.norm.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.projects.0.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.projects.0.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.projects.1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.projects.1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.projects.2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.projects.2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.projects.3.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.projects.3.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.resize_layers.0.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.resize_layers.0.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.resize_layers.1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.resize_layers.1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.resize_layers.3.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.resize_layers.3.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.layer1_rn.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.layer2_rn.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.layer3_rn.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.layer4_rn.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet1.out_conv.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet1.out_conv.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet1.resConfUnit1.conv1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet1.resConfUnit1.conv1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet1.resConfUnit1.conv2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet1.resConfUnit1.conv2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet1.resConfUnit2.conv1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet1.resConfUnit2.conv1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet1.resConfUnit2.conv2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet1.resConfUnit2.conv2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet2.out_conv.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet2.out_conv.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet2.resConfUnit1.conv1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet2.resConfUnit1.conv1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet2.resConfUnit1.conv2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet2.resConfUnit1.conv2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet2.resConfUnit2.conv1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet2.resConfUnit2.conv1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet2.resConfUnit2.conv2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet2.resConfUnit2.conv2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet3.out_conv.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet3.out_conv.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet3.resConfUnit1.conv1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet3.resConfUnit1.conv1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet3.resConfUnit1.conv2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet3.resConfUnit1.conv2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet3.resConfUnit2.conv1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet3.resConfUnit2.conv1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet3.resConfUnit2.conv2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet3.resConfUnit2.conv2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet4.out_conv.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet4.out_conv.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet4.resConfUnit1.conv1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet4.resConfUnit1.conv1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet4.resConfUnit1.conv2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet4.resConfUnit1.conv2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet4.resConfUnit2.conv1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet4.resConfUnit2.conv1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet4.resConfUnit2.conv2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.refinenet4.resConfUnit2.conv2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.output_conv1.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.output_conv1.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.output_conv2.0.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.output_conv2.0.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.output_conv2.2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.core.core.depth_head.scratch.output_conv2.2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.conv2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.conv2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.seed_bin_regressor._net.0.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.seed_bin_regressor._net.0.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.seed_bin_regressor._net.2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.seed_bin_regressor._net.2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.seed_projector._net.0.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.seed_projector._net.0.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.seed_projector._net.2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.seed_projector._net.2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.projectors.0._net.0.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.projectors.0._net.0.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.projectors.0._net.2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.projectors.0._net.2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.projectors.1._net.0.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.projectors.1._net.0.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.projectors.1._net.2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.projectors.1._net.2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.projectors.2._net.0.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.projectors.2._net.0.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.projectors.2._net.2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.projectors.2._net.2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.projectors.3._net.0.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.projectors.3._net.0.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.projectors.3._net.2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.projectors.3._net.2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.attractors.0._net.0.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.attractors.0._net.0.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.attractors.0._net.2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.attractors.0._net.2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.attractors.1._net.0.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.attractors.1._net.0.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.attractors.1._net.2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.attractors.1._net.2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.attractors.2._net.0.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.attractors.2._net.0.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.attractors.2._net.2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.attractors.2._net.2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.attractors.3._net.0.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.attractors.3._net.0.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.attractors.3._net.2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.attractors.3._net.2.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.conditional_log_binomial.mlp.0.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.conditional_log_binomial.mlp.0.bias
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.conditional_log_binomial.mlp.2.weight
+2024/03/15 03:55:33 - patchstitcher - INFO - training param: module.fine_branch.conditional_log_binomial.mlp.2.bias
+2024/03/15 03:57:49 - patchstitcher - INFO - Epoch: [01/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 2.039879322052002 - fine_loss: 2.039879322052002
+2024/03/15 03:59:40 - patchstitcher - INFO - Epoch: [01/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 3.776620626449585 - fine_loss: 3.776620626449585
+2024/03/15 04:01:30 - patchstitcher - INFO - Epoch: [01/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 2.1612205505371094 - fine_loss: 2.1612205505371094
+2024/03/15 04:03:20 - patchstitcher - INFO - Epoch: [01/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.3563077449798584 - fine_loss: 1.3563077449798584
+2024/03/15 04:06:31 - patchstitcher - INFO - Epoch: [02/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 2.1678900718688965 - fine_loss: 2.1678900718688965
+2024/03/15 04:08:25 - patchstitcher - INFO - Epoch: [02/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.8825774192810059 - fine_loss: 1.8825774192810059
+2024/03/15 04:10:14 - patchstitcher - INFO - Epoch: [02/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 2.350590467453003 - fine_loss: 2.350590467453003
+2024/03/15 04:12:06 - patchstitcher - INFO - Epoch: [02/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 2.691840648651123 - fine_loss: 2.691840648651123
+2024/03/15 04:13:51 - patchstitcher - INFO - Evaluation Summary: 
++----------+-----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+|    a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  |  rmse_log |   silog    |   sq_rel  |    see    |
++----------+-----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+| 0.707044 | 0.9293698 | 0.9801447 | 0.1927294 | 2.3443637 | 0.0782506 | 0.2331481 | 20.0879481 | 0.4492522 | 1.7012854 |
++----------+-----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+2024/03/15 04:15:48 - patchstitcher - INFO - Epoch: [03/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.2447803020477295 - fine_loss: 1.2447803020477295
+2024/03/15 04:17:37 - patchstitcher - INFO - Epoch: [03/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.6822900772094727 - fine_loss: 1.6822900772094727
+2024/03/15 04:19:22 - patchstitcher - INFO - Epoch: [03/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 2.7436625957489014 - fine_loss: 2.7436625957489014
+2024/03/15 04:21:15 - patchstitcher - INFO - Epoch: [03/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.9489283561706543 - fine_loss: 1.9489283561706543
+2024/03/15 04:24:21 - patchstitcher - INFO - Epoch: [04/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.5366265773773193 - fine_loss: 1.5366265773773193
+2024/03/15 04:26:10 - patchstitcher - INFO - Epoch: [04/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 2.0812580585479736 - fine_loss: 2.0812580585479736
+2024/03/15 04:28:00 - patchstitcher - INFO - Epoch: [04/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 2.318430185317993 - fine_loss: 2.318430185317993
+2024/03/15 04:29:48 - patchstitcher - INFO - Epoch: [04/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.638041615486145 - fine_loss: 1.638041615486145
+2024/03/15 04:31:27 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+----------+-----------+-----------+------------+-----------+----------+
+|     a1    |     a2    |     a3    |  abs_rel  |   rmse   |   log_10  |  rmse_log |   silog    |   sq_rel  |   see    |
++-----------+-----------+-----------+-----------+----------+-----------+-----------+------------+-----------+----------+
+| 0.7926732 | 0.9574633 | 0.9874803 | 0.1658911 | 2.033809 | 0.0653377 | 0.1971613 | 17.6386279 | 0.3764188 | 1.566062 |
++-----------+-----------+-----------+-----------+----------+-----------+-----------+------------+-----------+----------+
+2024/03/15 04:33:22 - patchstitcher - INFO - Epoch: [05/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.071550726890564 - fine_loss: 1.071550726890564
+2024/03/15 04:35:11 - patchstitcher - INFO - Epoch: [05/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.159848928451538 - fine_loss: 1.159848928451538
+2024/03/15 04:36:58 - patchstitcher - INFO - Epoch: [05/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.2986273765563965 - fine_loss: 1.2986273765563965
+2024/03/15 04:38:48 - patchstitcher - INFO - Epoch: [05/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.5721113681793213 - fine_loss: 1.5721113681793213
+2024/03/15 04:42:00 - patchstitcher - INFO - Epoch: [06/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.7645320892333984 - fine_loss: 1.7645320892333984
+2024/03/15 04:43:48 - patchstitcher - INFO - Epoch: [06/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.2818663120269775 - fine_loss: 1.2818663120269775
+2024/03/15 04:45:40 - patchstitcher - INFO - Epoch: [06/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.2445242404937744 - fine_loss: 1.2445242404937744
+2024/03/15 04:47:30 - patchstitcher - INFO - Epoch: [06/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.5368983745574951 - fine_loss: 1.5368983745574951
+2024/03/15 04:49:02 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  |  rmse_log |   silog    |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+| 0.8194143 | 0.9697637 | 0.9905658 | 0.1490125 | 1.8480574 | 0.0592408 | 0.1810736 | 15.8342003 | 0.3005681 | 1.3977808 |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+2024/03/15 04:50:58 - patchstitcher - INFO - Epoch: [07/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.4791369438171387 - fine_loss: 1.4791369438171387
+2024/03/15 04:52:44 - patchstitcher - INFO - Epoch: [07/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 2.1252331733703613 - fine_loss: 2.1252331733703613
+2024/03/15 04:54:32 - patchstitcher - INFO - Epoch: [07/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.84209406375885 - fine_loss: 1.84209406375885
+2024/03/15 04:56:25 - patchstitcher - INFO - Epoch: [07/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.1359673738479614 - fine_loss: 1.1359673738479614
+2024/03/15 04:59:38 - patchstitcher - INFO - Epoch: [08/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.5866280794143677 - fine_loss: 1.5866280794143677
+2024/03/15 05:01:29 - patchstitcher - INFO - Epoch: [08/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.3199617862701416 - fine_loss: 1.3199617862701416
+2024/03/15 05:03:15 - patchstitcher - INFO - Epoch: [08/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.6660882234573364 - fine_loss: 1.6660882234573364
+2024/03/15 05:05:05 - patchstitcher - INFO - Epoch: [08/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.0399880409240723 - fine_loss: 1.0399880409240723
+2024/03/15 05:06:40 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  |  rmse_log |   silog    |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+| 0.8804187 | 0.9831836 | 0.9948749 | 0.1118127 | 1.7537212 | 0.0498078 | 0.1550232 | 14.4210851 | 0.2352216 | 1.2980962 |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+2024/03/15 05:08:36 - patchstitcher - INFO - Epoch: [09/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.7554281949996948 - fine_loss: 1.7554281949996948
+2024/03/15 05:10:27 - patchstitcher - INFO - Epoch: [09/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 2.8572347164154053 - fine_loss: 2.8572347164154053
+2024/03/15 05:12:16 - patchstitcher - INFO - Epoch: [09/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.3657317161560059 - fine_loss: 1.3657317161560059
+2024/03/15 05:14:08 - patchstitcher - INFO - Epoch: [09/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.3460898399353027 - fine_loss: 1.3460898399353027
+2024/03/15 05:17:20 - patchstitcher - INFO - Epoch: [10/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.0736647844314575 - fine_loss: 1.0736647844314575
+2024/03/15 05:19:11 - patchstitcher - INFO - Epoch: [10/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.179059624671936 - fine_loss: 1.179059624671936
+2024/03/15 05:21:00 - patchstitcher - INFO - Epoch: [10/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.0112545490264893 - fine_loss: 1.0112545490264893
+2024/03/15 05:22:47 - patchstitcher - INFO - Epoch: [10/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.2453086376190186 - fine_loss: 1.2453086376190186
+2024/03/15 05:24:25 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  |  rmse_log |   silog    |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+| 0.8772274 | 0.9823961 | 0.9948375 | 0.1173125 | 1.7241426 | 0.0501591 | 0.1553792 | 14.1530364 | 0.2422748 | 1.3415729 |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+2024/03/15 05:26:18 - patchstitcher - INFO - Epoch: [11/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.306344747543335 - fine_loss: 1.306344747543335
+2024/03/15 05:28:16 - patchstitcher - INFO - Epoch: [11/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.348771572113037 - fine_loss: 1.348771572113037
+2024/03/15 05:30:06 - patchstitcher - INFO - Epoch: [11/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.549656629562378 - fine_loss: 1.549656629562378
+2024/03/15 05:31:57 - patchstitcher - INFO - Epoch: [11/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.4452790021896362 - fine_loss: 1.4452790021896362
+2024/03/15 05:35:10 - patchstitcher - INFO - Epoch: [12/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.1077752113342285 - fine_loss: 1.1077752113342285
+2024/03/15 05:37:01 - patchstitcher - INFO - Epoch: [12/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.8956596255302429 - fine_loss: 0.8956596255302429
+2024/03/15 05:38:52 - patchstitcher - INFO - Epoch: [12/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.9720367789268494 - fine_loss: 0.9720367789268494
+2024/03/15 05:40:41 - patchstitcher - INFO - Epoch: [12/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.4826208353042603 - fine_loss: 1.4826208353042603
+2024/03/15 05:42:16 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+----------+-----------+-----------+------------+----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |   rmse   |   log_10  |  rmse_log |   silog    |  sq_rel  |    see    |
++-----------+-----------+-----------+-----------+----------+-----------+-----------+------------+----------+-----------+
+| 0.8740682 | 0.9844725 | 0.9957269 | 0.1142447 | 1.696142 | 0.0509766 | 0.1547242 | 13.9800131 | 0.237403 | 1.2716073 |
++-----------+-----------+-----------+-----------+----------+-----------+-----------+------------+----------+-----------+
+2024/03/15 05:44:13 - patchstitcher - INFO - Epoch: [13/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.7906665802001953 - fine_loss: 1.7906665802001953
+2024/03/15 05:46:06 - patchstitcher - INFO - Epoch: [13/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.7277212142944336 - fine_loss: 1.7277212142944336
+2024/03/15 05:48:00 - patchstitcher - INFO - Epoch: [13/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.1345900297164917 - fine_loss: 1.1345900297164917
+2024/03/15 05:49:53 - patchstitcher - INFO - Epoch: [13/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.680286169052124 - fine_loss: 0.680286169052124
+2024/03/15 05:53:05 - patchstitcher - INFO - Epoch: [14/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.0135771036148071 - fine_loss: 1.0135771036148071
+2024/03/15 05:54:56 - patchstitcher - INFO - Epoch: [14/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.1816802024841309 - fine_loss: 1.1816802024841309
+2024/03/15 05:56:44 - patchstitcher - INFO - Epoch: [14/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.3476241827011108 - fine_loss: 1.3476241827011108
+2024/03/15 05:58:33 - patchstitcher - INFO - Epoch: [14/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.6280028223991394 - fine_loss: 0.6280028223991394
+2024/03/15 06:00:11 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+-----------+----------+------------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  | rmse_log |   silog    |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+-----------+-----------+----------+------------+-----------+-----------+
+| 0.9147314 | 0.9859354 | 0.9949076 | 0.1007045 | 1.6106567 | 0.0434999 | 0.138901 | 13.0318626 | 0.2056279 | 1.2140529 |
++-----------+-----------+-----------+-----------+-----------+-----------+----------+------------+-----------+-----------+
+2024/03/15 06:02:04 - patchstitcher - INFO - Epoch: [15/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.8376606702804565 - fine_loss: 0.8376606702804565
+2024/03/15 06:03:57 - patchstitcher - INFO - Epoch: [15/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.03225576877594 - fine_loss: 1.03225576877594
+2024/03/15 06:05:44 - patchstitcher - INFO - Epoch: [15/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.9883253574371338 - fine_loss: 0.9883253574371338
+2024/03/15 06:07:36 - patchstitcher - INFO - Epoch: [15/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.262385368347168 - fine_loss: 1.262385368347168
+2024/03/15 06:10:46 - patchstitcher - INFO - Epoch: [16/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.1695902347564697 - fine_loss: 1.1695902347564697
+2024/03/15 06:12:36 - patchstitcher - INFO - Epoch: [16/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 2.1688151359558105 - fine_loss: 2.1688151359558105
+2024/03/15 06:14:24 - patchstitcher - INFO - Epoch: [16/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.3791565895080566 - fine_loss: 1.3791565895080566
+2024/03/15 06:16:12 - patchstitcher - INFO - Epoch: [16/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.2718651294708252 - fine_loss: 1.2718651294708252
+2024/03/15 06:17:50 - patchstitcher - INFO - Evaluation Summary: 
++----------+-----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+|    a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  |  rmse_log |   silog    |   sq_rel  |    see    |
++----------+-----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+| 0.917846 | 0.9849823 | 0.9948954 | 0.0979613 | 1.5791011 | 0.0433261 | 0.1380226 | 12.8257169 | 0.1883265 | 1.1684257 |
++----------+-----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+2024/03/15 06:19:42 - patchstitcher - INFO - Epoch: [17/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.0557522773742676 - fine_loss: 1.0557522773742676
+2024/03/15 06:21:33 - patchstitcher - INFO - Epoch: [17/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.6954542398452759 - fine_loss: 0.6954542398452759
+2024/03/15 06:23:20 - patchstitcher - INFO - Epoch: [17/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.203284740447998 - fine_loss: 1.203284740447998
+2024/03/15 06:25:09 - patchstitcher - INFO - Epoch: [17/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.2890739440917969 - fine_loss: 1.2890739440917969
+2024/03/15 06:28:25 - patchstitcher - INFO - Epoch: [18/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.8028295040130615 - fine_loss: 0.8028295040130615
+2024/03/15 06:30:13 - patchstitcher - INFO - Epoch: [18/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.499220609664917 - fine_loss: 0.499220609664917
+2024/03/15 06:32:01 - patchstitcher - INFO - Epoch: [18/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.8515260219573975 - fine_loss: 0.8515260219573975
+2024/03/15 06:33:51 - patchstitcher - INFO - Epoch: [18/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.953697919845581 - fine_loss: 0.953697919845581
+2024/03/15 06:35:27 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  |  rmse_log |   silog    |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+| 0.9318894 | 0.9858092 | 0.9957183 | 0.0923063 | 1.5112557 | 0.0394818 | 0.1269675 | 11.6301649 | 0.1761516 | 1.1147971 |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+2024/03/15 06:37:24 - patchstitcher - INFO - Epoch: [19/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.8879871368408203 - fine_loss: 0.8879871368408203
+2024/03/15 06:39:14 - patchstitcher - INFO - Epoch: [19/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.4138840436935425 - fine_loss: 1.4138840436935425
+2024/03/15 06:41:05 - patchstitcher - INFO - Epoch: [19/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.3911192417144775 - fine_loss: 1.3911192417144775
+2024/03/15 06:42:59 - patchstitcher - INFO - Epoch: [19/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.9037826061248779 - fine_loss: 0.9037826061248779
+2024/03/15 06:46:06 - patchstitcher - INFO - Epoch: [20/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.7059022784233093 - fine_loss: 0.7059022784233093
+2024/03/15 06:47:58 - patchstitcher - INFO - Epoch: [20/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.8616353273391724 - fine_loss: 0.8616353273391724
+2024/03/15 06:49:51 - patchstitcher - INFO - Epoch: [20/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.8395438194274902 - fine_loss: 0.8395438194274902
+2024/03/15 06:51:43 - patchstitcher - INFO - Epoch: [20/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.6362200379371643 - fine_loss: 0.6362200379371643
+2024/03/15 06:53:21 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+-----------+----------+------------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  | rmse_log |   silog    |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+-----------+-----------+----------+------------+-----------+-----------+
+| 0.9486918 | 0.9883879 | 0.9965515 | 0.0802352 | 1.4414517 | 0.0349744 | 0.116316 | 10.9957016 | 0.1575956 | 1.0969994 |
++-----------+-----------+-----------+-----------+-----------+-----------+----------+------------+-----------+-----------+
+2024/03/15 06:55:18 - patchstitcher - INFO - Epoch: [21/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.6189630627632141 - fine_loss: 0.6189630627632141
+2024/03/15 06:57:11 - patchstitcher - INFO - Epoch: [21/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.1719452142715454 - fine_loss: 1.1719452142715454
+2024/03/15 06:58:55 - patchstitcher - INFO - Epoch: [21/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.142961025238037 - fine_loss: 1.142961025238037
+2024/03/15 07:00:45 - patchstitcher - INFO - Epoch: [21/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.719948649406433 - fine_loss: 1.719948649406433
+2024/03/15 07:03:58 - patchstitcher - INFO - Epoch: [22/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 0.6470488905906677 - fine_loss: 0.6470488905906677
+2024/03/15 07:05:49 - patchstitcher - INFO - Epoch: [22/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.5520279407501221 - fine_loss: 0.5520279407501221
+2024/03/15 07:07:38 - patchstitcher - INFO - Epoch: [22/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.8810967803001404 - fine_loss: 0.8810967803001404
+2024/03/15 07:09:32 - patchstitcher - INFO - Epoch: [22/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.6827142238616943 - fine_loss: 0.6827142238616943
+2024/03/15 07:11:07 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  |  rmse_log |   silog   |   sq_rel  |   see    |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+----------+
+| 0.9523656 | 0.9892937 | 0.9967417 | 0.0767006 | 1.4133022 | 0.0333895 | 0.1125023 | 10.666611 | 0.1523504 | 1.061902 |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+-----------+----------+
+2024/03/15 07:13:02 - patchstitcher - INFO - Epoch: [23/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.8002086877822876 - fine_loss: 1.8002086877822876
+2024/03/15 07:14:51 - patchstitcher - INFO - Epoch: [23/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 0.5043245553970337 - fine_loss: 0.5043245553970337
+2024/03/15 07:16:39 - patchstitcher - INFO - Epoch: [23/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 1.6025413274765015 - fine_loss: 1.6025413274765015
+2024/03/15 07:18:29 - patchstitcher - INFO - Epoch: [23/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 1.3183393478393555 - fine_loss: 1.3183393478393555
+2024/03/15 07:21:41 - patchstitcher - INFO - Epoch: [24/24] - Step: [00100/00475] - Time: [1/1] - Total Loss: 1.6571695804595947 - fine_loss: 1.6571695804595947
+2024/03/15 07:23:30 - patchstitcher - INFO - Epoch: [24/24] - Step: [00200/00475] - Time: [1/1] - Total Loss: 1.0306520462036133 - fine_loss: 1.0306520462036133
+2024/03/15 07:25:19 - patchstitcher - INFO - Epoch: [24/24] - Step: [00300/00475] - Time: [1/1] - Total Loss: 0.8030037879943848 - fine_loss: 0.8030037879943848
+2024/03/15 07:27:09 - patchstitcher - INFO - Epoch: [24/24] - Step: [00400/00475] - Time: [1/1] - Total Loss: 0.6139640808105469 - fine_loss: 0.6139640808105469
+2024/03/15 07:28:49 - patchstitcher - INFO - Evaluation Summary: 
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+|     a1    |     a2    |     a3    |  abs_rel  |    rmse   |   log_10  |  rmse_log |   silog    |   sq_rel  |    see    |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+| 0.9531358 | 0.9897053 | 0.9967571 | 0.0759499 | 1.4041272 | 0.0327699 | 0.1107659 | 10.5243982 | 0.1508702 | 1.0635976 |
++-----------+-----------+-----------+-----------+-----------+-----------+-----------+------------+-----------+-----------+
+2024/03/15 07:28:49 - patchstitcher - INFO - Saving ckp, but use the inner get_save_dict fuction to get model_dict
+2024/03/15 07:28:49 - patchstitcher - INFO - For saving space. Would you like to save base model several times? :>
+2024/03/15 07:28:49 - patchstitcher - INFO - save checkpoint_24.pth at ./work_dir/depthanything_vits_u4k/fine_pretrain

depthanything_vits_u4k/fine_pretrain/checkpoint_24.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:83e1263ca84ae622c57b1cdfc1e1e2b548ad13d3e1d2b143193a6d5efcc5fc92
+size 300162730

depthanything_vits_u4k/fine_pretrain/config.py

@@ -0,0 +1,314 @@
+collect_input_args = [
+    'image_lr',
+    'crops_image_hr',
+    'depth_gt',
+    'crop_depths',
+    'bboxs',
+    'image_hr',
+]
+convert_syncbn = True
+debug = False
+env_cfg = dict(
+    cudnn_benchmark=True,
+    dist_cfg=dict(backend='nccl'),
+    mp_cfg=dict(mp_start_method='forkserver'))
+find_unused_parameters = True
+general_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        dataset_name='', gt_dir=None, rgb_image_dir='', type='ImageDataset'),
+    num_workers=2)
+launcher = 'pytorch'
+log_name = 'fine_pretrain'
+max_depth = 80
+min_depth = 0.001
+model = dict(
+    coarse_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        depth_anything=True,
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            392,
+            518,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='vits',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource='local::./work_dir/DepthAnything_vits.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='DA-ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    fine_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        depth_anything=True,
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            392,
+            518,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='vits',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource='local::./work_dir/DepthAnything_vits.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='DA-ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    max_depth=80,
+    min_depth=0.001,
+    patch_process_shape=(
+        392,
+        518,
+    ),
+    sigloss=dict(type='SILogLoss'),
+    target='fine',
+    type='BaselinePretrain')
+optim_wrapper = dict(
+    clip_grad=dict(max_norm=0.1, norm_type=2, type='norm'),
+    optimizer=dict(lr=4e-06, type='AdamW', weight_decay=0.01),
+    paramwise_cfg=dict(bypass_duplicate=True, custom_keys=dict()))
+param_scheduler = dict(
+    base_momentum=0.85,
+    cycle_momentum=True,
+    div_factor=1,
+    final_div_factor=10000,
+    max_momentum=0.95,
+    pct_start=0.5,
+    three_phase=False)
+project = 'patchfusion'
+resume = False
+tags = [
+    'fine',
+    'da',
+    'vits',
+]
+test_in_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+test_out_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test_out.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+train_cfg = dict(
+    eval_start=0,
+    log_interval=100,
+    max_epochs=24,
+    save_checkpoint_interval=24,
+    train_log_img_interval=100,
+    val_interval=2,
+    val_log_img_interval=50,
+    val_type='epoch_base')
+train_dataloader = dict(
+    batch_size=4,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='train',
+        resize_mode='depth-anything',
+        split='./data/u4k/splits/train.txt',
+        transform_cfg=dict(
+            degree=1.0,
+            network_process_size=[
+                392,
+                518,
+            ],
+            random_crop=True,
+            random_crop_size=(
+                540,
+                960,
+            )),
+        type='UnrealStereo4kDataset'),
+    num_workers=4)
+val_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        resize_mode='depth-anything',
+        split='./data/u4k/splits/val.txt',
+        transform_cfg=dict(
+            degree=1.0,
+            network_process_size=[
+                392,
+                518,
+            ],
+            random_crop_size=(
+                540,
+                960,
+            )),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+work_dir = './work_dir/depthanything_vits_u4k/fine_pretrain'
+zoe_depth_config = dict(
+    attractor_alpha=1000,
+    attractor_gamma=2,
+    attractor_kind='mean',
+    attractor_type='inv',
+    aug=True,
+    bin_centers_type='softplus',
+    bin_embedding_dim=128,
+    clip_grad=0.1,
+    dataset='nyu',
+    depth_anything=True,
+    distributed=True,
+    do_resize=False,
+    force_keep_ar=True,
+    freeze_midas_bn=True,
+    gpu='NULL',
+    img_size=[
+        392,
+        518,
+    ],
+    inverse_midas=False,
+    log_images_every=0.1,
+    max_depth=80,
+    max_temp=50.0,
+    max_translation=100,
+    memory_efficient=True,
+    midas_model_type='vits',
+    min_depth=0.001,
+    min_temp=0.0212,
+    model='zoedepth',
+    n_attractors=[
+        16,
+        8,
+        4,
+        1,
+    ],
+    n_bins=64,
+    name='ZoeDepth',
+    notes='',
+    output_distribution='logbinomial',
+    prefetch=False,
+    pretrained_resource='local::./work_dir/DepthAnything_vits.pt',
+    print_losses=False,
+    project='ZoeDepth',
+    random_crop=False,
+    random_translate=False,
+    root='.',
+    save_dir='',
+    shared_dict='NULL',
+    tags='',
+    train_midas=True,
+    translate_prob=0.2,
+    type='DA-ZoeDepth',
+    uid='NULL',
+    use_amp=False,
+    use_pretrained_midas=True,
+    use_shared_dict=False,
+    validate_every=0.25,
+    version_name='v1',
+    workers=16)

depthanything_vits_u4k/patchfusion/checkpoint_16.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:12c10292f758cc38f45bd2dad11240190d440c8416b5834180642253f0ea93b9
+size 205127853

depthanything_vits_u4k/patchfusion/config.py

@@ -0,0 +1,341 @@
+collect_input_args = [
+    'image_lr',
+    'crops_image_hr',
+    'depth_gt',
+    'crop_depths',
+    'bboxs',
+    'image_hr',
+]
+convert_syncbn = True
+debug = False
+env_cfg = dict(
+    cudnn_benchmark=True,
+    dist_cfg=dict(backend='nccl'),
+    mp_cfg=dict(mp_start_method='forkserver'))
+find_unused_parameters = True
+general_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        dataset_name='', gt_dir=None, rgb_image_dir='', type='ImageDataset'),
+    num_workers=2)
+launcher = 'pytorch'
+log_name = 'patchfusion'
+max_depth = 80
+min_depth = 0.001
+model = dict(
+    coarse_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        depth_anything=True,
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            392,
+            518,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='vits',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource='local::./work_dir/DepthAnything_vits.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='DA-ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    fine_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        depth_anything=True,
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            392,
+            518,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='vits',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource='local::./work_dir/DepthAnything_vits.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='DA-ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    guided_fusion=dict(
+        g2l=True,
+        in_channels=[
+            32,
+            64,
+            64,
+            64,
+            64,
+            64,
+        ],
+        n_channels=5,
+        num_patches=[
+            203056,
+            66304,
+            16576,
+            4144,
+            1036,
+            266,
+        ],
+        patch_process_shape=(
+            392,
+            518,
+        ),
+        type='GuidedFusionPatchFusion'),
+    max_depth=80,
+    min_depth=0.001,
+    patch_process_shape=(
+        392,
+        518,
+    ),
+    pretrain_model=[
+        './work_dir/depthanything_vits_u4k/coarse_pretrain/checkpoint_24.pth',
+        './work_dir/depthanything_vits_u4k/fine_pretrain/checkpoint_24.pth',
+    ],
+    sigloss=dict(type='SILogLoss'),
+    type='PatchFusion')
+optim_wrapper = dict(
+    clip_grad=dict(max_norm=0.1, norm_type=2, type='norm'),
+    optimizer=dict(lr=0.0001, type='AdamW', weight_decay=0.001),
+    paramwise_cfg=dict(bypass_duplicate=True, custom_keys=dict()))
+param_scheduler = dict(
+    base_momentum=0.85,
+    cycle_momentum=True,
+    div_factor=10,
+    final_div_factor=10000,
+    max_momentum=0.95,
+    pct_start=0.25,
+    three_phase=False)
+project = 'patchfusion'
+resume = False
+tags = [
+    'patchfusion',
+    'da',
+    'vits',
+]
+test_in_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+test_out_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test_out.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+train_cfg = dict(
+    eval_start=0,
+    log_interval=100,
+    max_epochs=16,
+    save_checkpoint_interval=16,
+    train_log_img_interval=500,
+    val_interval=2,
+    val_log_img_interval=50,
+    val_type='epoch_base')
+train_dataloader = dict(
+    batch_size=4,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='train',
+        resize_mode='depth-anything',
+        split='./data/u4k/splits/train.txt',
+        transform_cfg=dict(
+            degree=1.0,
+            network_process_size=[
+                392,
+                518,
+            ],
+            random_crop=True,
+            random_crop_size=(
+                540,
+                960,
+            )),
+        type='UnrealStereo4kDataset'),
+    num_workers=4)
+val_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        resize_mode='depth-anything',
+        split='./data/u4k/splits/val.txt',
+        transform_cfg=dict(
+            degree=1.0,
+            network_process_size=[
+                392,
+                518,
+            ],
+            random_crop_size=(
+                540,
+                960,
+            )),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+work_dir = './work_dir/depthanything_vits_u4k/patchfusion'
+zoe_depth_config = dict(
+    attractor_alpha=1000,
+    attractor_gamma=2,
+    attractor_kind='mean',
+    attractor_type='inv',
+    aug=True,
+    bin_centers_type='softplus',
+    bin_embedding_dim=128,
+    clip_grad=0.1,
+    dataset='nyu',
+    depth_anything=True,
+    distributed=True,
+    do_resize=False,
+    force_keep_ar=True,
+    freeze_midas_bn=True,
+    gpu='NULL',
+    img_size=[
+        392,
+        518,
+    ],
+    inverse_midas=False,
+    log_images_every=0.1,
+    max_depth=80,
+    max_temp=50.0,
+    max_translation=100,
+    memory_efficient=True,
+    midas_model_type='vits',
+    min_depth=0.001,
+    min_temp=0.0212,
+    model='zoedepth',
+    n_attractors=[
+        16,
+        8,
+        4,
+        1,
+    ],
+    n_bins=64,
+    name='ZoeDepth',
+    notes='',
+    output_distribution='logbinomial',
+    prefetch=False,
+    pretrained_resource='local::./work_dir/DepthAnything_vits.pt',
+    print_losses=False,
+    project='ZoeDepth',
+    random_crop=False,
+    random_translate=False,
+    root='.',
+    save_dir='',
+    shared_dict='NULL',
+    tags='',
+    train_midas=True,
+    translate_prob=0.2,
+    type='DA-ZoeDepth',
+    uid='NULL',
+    use_amp=False,
+    use_pretrained_midas=True,
+    use_shared_dict=False,
+    validate_every=0.25,
+    version_name='v1',
+    workers=16)

zoedepth_u4k/coarse_pretrain/checkpoint_24.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b241f5a664d9b8b27f138d677ca56a9b1629838c4529cced25d00043daa85950
+size 4184807605

zoedepth_u4k/coarse_pretrain/config.py

@@ -0,0 +1,307 @@
+collect_input_args = [
+    'image_lr',
+    'crops_image_hr',
+    'depth_gt',
+    'crop_depths',
+    'bboxs',
+    'image_hr',
+]
+convert_syncbn = True
+debug = True
+env_cfg = dict(
+    cudnn_benchmark=True,
+    dist_cfg=dict(backend='nccl'),
+    mp_cfg=dict(mp_start_method='forkserver'))
+find_unused_parameters = True
+general_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        dataset_name='', gt_dir=None, rgb_image_dir='', type='ImageDataset'),
+    num_workers=2)
+launcher = 'pytorch'
+log_name = 'coarse_pretrain'
+max_depth = 80
+min_depth = 0.001
+model = dict(
+    coarse_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            384,
+            512,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='DPT_BEiT_L_384',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource=
+        'local::./work_dir/ZoeDepthv1_30-Dec_16-29-4e2bc436e4e1_best.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    fine_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            384,
+            512,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='DPT_BEiT_L_384',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource=
+        'local::./work_dir/ZoeDepthv1_30-Dec_16-29-4e2bc436e4e1_best.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    max_depth=80,
+    min_depth=0.001,
+    sigloss=dict(type='SILogLoss'),
+    target='coarse',
+    type='BaselinePretrain')
+optim_wrapper = dict(
+    clip_grad=dict(max_norm=0.1, norm_type=2, type='norm'),
+    optimizer=dict(lr=0.0002, type='AdamW', weight_decay=0.01),
+    paramwise_cfg=dict(
+        bypass_duplicate=True,
+        custom_keys=dict(
+            {'coarse_branch.core': dict(decay_mult=1.0, lr_mult=0.1)})))
+param_scheduler = dict(
+    base_momentum=0.85,
+    cycle_momentum=True,
+    div_factor=1,
+    final_div_factor=10000,
+    max_momentum=0.95,
+    pct_start=0.5,
+    three_phase=False)
+project = 'patchfusion'
+resume = False
+tags = [
+    'pcoarse',
+]
+test_in_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+test_out_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test_out.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+train_cfg = dict(
+    eval_start=0,
+    log_interval=100,
+    max_epochs=24,
+    save_checkpoint_interval=24,
+    train_log_img_interval=100,
+    val_interval=2,
+    val_log_img_interval=50,
+    val_type='epoch_base')
+train_dataloader = dict(
+    batch_size=4,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='train',
+        split='./data/u4k/splits/train.txt',
+        transform_cfg=dict(
+            degree=1.0,
+            network_process_size=[
+                384,
+                512,
+            ],
+            random_crop=True,
+            random_crop_size=(
+                540,
+                960,
+            )),
+        type='UnrealStereo4kDataset'),
+    num_workers=4)
+val_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/val.txt',
+        transform_cfg=dict(
+            network_process_size=[
+                384,
+                512,
+            ], random_crop_size=(
+                540,
+                960,
+            )),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+work_dir = './work_dir/coarse_pretrain'
+zoe_depth_config = dict(
+    attractor_alpha=1000,
+    attractor_gamma=2,
+    attractor_kind='mean',
+    attractor_type='inv',
+    aug=True,
+    bin_centers_type='softplus',
+    bin_embedding_dim=128,
+    clip_grad=0.1,
+    dataset='nyu',
+    distributed=True,
+    do_resize=False,
+    force_keep_ar=True,
+    freeze_midas_bn=True,
+    gpu='NULL',
+    img_size=[
+        384,
+        512,
+    ],
+    inverse_midas=False,
+    log_images_every=0.1,
+    max_depth=80,
+    max_temp=50.0,
+    max_translation=100,
+    memory_efficient=True,
+    midas_model_type='DPT_BEiT_L_384',
+    min_depth=0.001,
+    min_temp=0.0212,
+    model='zoedepth',
+    n_attractors=[
+        16,
+        8,
+        4,
+        1,
+    ],
+    n_bins=64,
+    name='ZoeDepth',
+    notes='',
+    output_distribution='logbinomial',
+    prefetch=False,
+    pretrained_resource=
+    'local::./work_dir/ZoeDepthv1_30-Dec_16-29-4e2bc436e4e1_best.pt',
+    print_losses=False,
+    project='ZoeDepth',
+    random_crop=False,
+    random_translate=False,
+    root='.',
+    save_dir='',
+    shared_dict='NULL',
+    tags='',
+    train_midas=True,
+    translate_prob=0.2,
+    type='ZoeDepth',
+    uid='NULL',
+    use_amp=False,
+    use_pretrained_midas=True,
+    use_shared_dict=False,
+    validate_every=0.25,
+    version_name='v1',
+    workers=16)

zoedepth_u4k/fine_pretrain/checkpoint_24.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8915b17cf436ed3e1e4cfe9c3eecdfb806aa5dc0f66924de6e785f4a16c431e2
+size 4184807669

zoedepth_u4k/fine_pretrain/config.py

@@ -0,0 +1,307 @@
+collect_input_args = [
+    'image_lr',
+    'crops_image_hr',
+    'depth_gt',
+    'crop_depths',
+    'bboxs',
+    'image_hr',
+]
+convert_syncbn = True
+debug = True
+env_cfg = dict(
+    cudnn_benchmark=True,
+    dist_cfg=dict(backend='nccl'),
+    mp_cfg=dict(mp_start_method='forkserver'))
+find_unused_parameters = True
+general_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        dataset_name='', gt_dir=None, rgb_image_dir='', type='ImageDataset'),
+    num_workers=2)
+launcher = 'pytorch'
+log_name = 'fine_pretrain'
+max_depth = 80
+min_depth = 0.001
+model = dict(
+    coarse_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            384,
+            512,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='DPT_BEiT_L_384',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource=
+        'local::./work_dir/ZoeDepthv1_30-Dec_16-29-4e2bc436e4e1_best.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    fine_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            384,
+            512,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='DPT_BEiT_L_384',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource=
+        'local::./work_dir/ZoeDepthv1_30-Dec_16-29-4e2bc436e4e1_best.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    max_depth=80,
+    min_depth=0.001,
+    sigloss=dict(type='SILogLoss'),
+    target='fine',
+    type='BaselinePretrain')
+optim_wrapper = dict(
+    clip_grad=dict(max_norm=0.1, norm_type=2, type='norm'),
+    optimizer=dict(lr=0.0002, type='AdamW', weight_decay=0.01),
+    paramwise_cfg=dict(
+        bypass_duplicate=True,
+        custom_keys=dict(
+            {'fine_branch.core': dict(decay_mult=1.0, lr_mult=0.1)})))
+param_scheduler = dict(
+    base_momentum=0.85,
+    cycle_momentum=True,
+    div_factor=1,
+    final_div_factor=10000,
+    max_momentum=0.95,
+    pct_start=0.5,
+    three_phase=False)
+project = 'patchfusion'
+resume = False
+tags = [
+    'fine',
+]
+test_in_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+test_out_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test_out.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+train_cfg = dict(
+    eval_start=0,
+    log_interval=100,
+    max_epochs=24,
+    save_checkpoint_interval=24,
+    train_log_img_interval=100,
+    val_interval=2,
+    val_log_img_interval=50,
+    val_type='epoch_base')
+train_dataloader = dict(
+    batch_size=4,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='train',
+        split='./data/u4k/splits/train.txt',
+        transform_cfg=dict(
+            degree=1.0,
+            network_process_size=[
+                384,
+                512,
+            ],
+            random_crop=True,
+            random_crop_size=(
+                540,
+                960,
+            )),
+        type='UnrealStereo4kDataset'),
+    num_workers=4)
+val_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/val.txt',
+        transform_cfg=dict(
+            network_process_size=[
+                384,
+                512,
+            ], random_crop_size=(
+                540,
+                960,
+            )),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+work_dir = './work_dir/fine_pretrain'
+zoe_depth_config = dict(
+    attractor_alpha=1000,
+    attractor_gamma=2,
+    attractor_kind='mean',
+    attractor_type='inv',
+    aug=True,
+    bin_centers_type='softplus',
+    bin_embedding_dim=128,
+    clip_grad=0.1,
+    dataset='nyu',
+    distributed=True,
+    do_resize=False,
+    force_keep_ar=True,
+    freeze_midas_bn=True,
+    gpu='NULL',
+    img_size=[
+        384,
+        512,
+    ],
+    inverse_midas=False,
+    log_images_every=0.1,
+    max_depth=80,
+    max_temp=50.0,
+    max_translation=100,
+    memory_efficient=True,
+    midas_model_type='DPT_BEiT_L_384',
+    min_depth=0.001,
+    min_temp=0.0212,
+    model='zoedepth',
+    n_attractors=[
+        16,
+        8,
+        4,
+        1,
+    ],
+    n_bins=64,
+    name='ZoeDepth',
+    notes='',
+    output_distribution='logbinomial',
+    prefetch=False,
+    pretrained_resource=
+    'local::./work_dir/ZoeDepthv1_30-Dec_16-29-4e2bc436e4e1_best.pt',
+    print_losses=False,
+    project='ZoeDepth',
+    random_crop=False,
+    random_translate=False,
+    root='.',
+    save_dir='',
+    shared_dict='NULL',
+    tags='',
+    train_midas=True,
+    translate_prob=0.2,
+    type='ZoeDepth',
+    uid='NULL',
+    use_amp=False,
+    use_pretrained_midas=True,
+    use_shared_dict=False,
+    validate_every=0.25,
+    version_name='v1',
+    workers=16)

zoedepth_u4k/patchfusion/checkpoint_16.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e414483120fa6b95ced56e55b3cb5e711b076f9cc1f62a5c54d1edecc5c90fab
+size 1055616493

zoedepth_u4k/patchfusion/config.py

@@ -0,0 +1,305 @@
+collect_input_args = [
+    'image_lr',
+    'crops_image_hr',
+    'depth_gt',
+    'crop_depths',
+    'bboxs',
+    'image_hr',
+]
+convert_syncbn = True
+debug = False
+env_cfg = dict(
+    cudnn_benchmark=True,
+    dist_cfg=dict(backend='nccl'),
+    mp_cfg=dict(mp_start_method='forkserver'))
+find_unused_parameters = True
+general_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        dataset_name='', gt_dir=None, rgb_image_dir='', type='ImageDataset'),
+    num_workers=2)
+launcher = 'pytorch'
+log_name = 'patchfusion'
+max_depth = 80
+min_depth = 0.001
+model = dict(
+    coarse_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            384,
+            512,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='DPT_BEiT_L_384',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource='local::./work_dir/ZoeDepthv1.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    fine_branch=dict(
+        attractor_alpha=1000,
+        attractor_gamma=2,
+        attractor_kind='mean',
+        attractor_type='inv',
+        aug=True,
+        bin_centers_type='softplus',
+        bin_embedding_dim=128,
+        clip_grad=0.1,
+        dataset='nyu',
+        distributed=True,
+        do_resize=False,
+        force_keep_ar=True,
+        freeze_midas_bn=True,
+        gpu='NULL',
+        img_size=[
+            384,
+            512,
+        ],
+        inverse_midas=False,
+        log_images_every=0.1,
+        max_depth=80,
+        max_temp=50.0,
+        max_translation=100,
+        memory_efficient=True,
+        midas_model_type='DPT_BEiT_L_384',
+        min_depth=0.001,
+        min_temp=0.0212,
+        model='zoedepth',
+        n_attractors=[
+            16,
+            8,
+            4,
+            1,
+        ],
+        n_bins=64,
+        name='ZoeDepth',
+        notes='',
+        output_distribution='logbinomial',
+        prefetch=False,
+        pretrained_resource='local::./work_dir/ZoeDepthv1.pt',
+        print_losses=False,
+        project='ZoeDepth',
+        random_crop=False,
+        random_translate=False,
+        root='.',
+        save_dir='',
+        shared_dict='NULL',
+        tags='',
+        train_midas=True,
+        translate_prob=0.2,
+        type='ZoeDepth',
+        uid='NULL',
+        use_amp=False,
+        use_pretrained_midas=True,
+        use_shared_dict=False,
+        validate_every=0.25,
+        version_name='v1',
+        workers=16),
+    guided_fusion=dict(g2l=True, n_channels=5, type='GuidedFusionPatchFusion'),
+    max_depth=80,
+    min_depth=0.001,
+    pretrain_model=[
+        './work_dir/coarse_pretrain/checkpoint_24.pth',
+        './work_dir/fine_pretrain/checkpoint_24.pth',
+    ],
+    sigloss=dict(type='SILogLoss'),
+    type='PatchFusion')
+optim_wrapper = dict(
+    clip_grad=dict(max_norm=0.1, norm_type=2, type='norm'),
+    optimizer=dict(lr=0.0001, type='AdamW', weight_decay=0.001),
+    paramwise_cfg=dict(bypass_duplicate=True, custom_keys=dict()))
+param_scheduler = dict(
+    base_momentum=0.85,
+    cycle_momentum=True,
+    div_factor=10,
+    final_div_factor=10000,
+    max_momentum=0.95,
+    pct_start=0.25,
+    three_phase=False)
+project = 'patchfusion'
+resume = False
+tags = [
+    'patchfusion',
+]
+test_in_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+test_out_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/test_out.txt',
+        transform_cfg=dict(network_process_size=[
+            384,
+            512,
+        ]),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+train_cfg = dict(
+    eval_start=0,
+    log_interval=100,
+    max_epochs=16,
+    save_checkpoint_interval=16,
+    train_log_img_interval=500,
+    val_interval=2,
+    val_log_img_interval=10,
+    val_type='epoch_base')
+train_dataloader = dict(
+    batch_size=4,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='train',
+        split='./data/u4k/splits/train.txt',
+        transform_cfg=dict(
+            degree=1.0,
+            network_process_size=[
+                384,
+                512,
+            ],
+            random_crop=True,
+            random_crop_size=(
+                540,
+                960,
+            )),
+        type='UnrealStereo4kDataset'),
+    num_workers=4)
+val_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        data_root='./data/u4k',
+        max_depth=80,
+        min_depth=0.001,
+        mode='infer',
+        split='./data/u4k/splits/val.txt',
+        transform_cfg=dict(
+            network_process_size=[
+                384,
+                512,
+            ], random_crop_size=(
+                540,
+                960,
+            )),
+        type='UnrealStereo4kDataset'),
+    num_workers=2)
+work_dir = './work_dir/patchfusion'
+zoe_depth_config = dict(
+    attractor_alpha=1000,
+    attractor_gamma=2,
+    attractor_kind='mean',
+    attractor_type='inv',
+    aug=True,
+    bin_centers_type='softplus',
+    bin_embedding_dim=128,
+    clip_grad=0.1,
+    dataset='nyu',
+    distributed=True,
+    do_resize=False,
+    force_keep_ar=True,
+    freeze_midas_bn=True,
+    gpu='NULL',
+    img_size=[
+        384,
+        512,
+    ],
+    inverse_midas=False,
+    log_images_every=0.1,
+    max_depth=80,
+    max_temp=50.0,
+    max_translation=100,
+    memory_efficient=True,
+    midas_model_type='DPT_BEiT_L_384',
+    min_depth=0.001,
+    min_temp=0.0212,
+    model='zoedepth',
+    n_attractors=[
+        16,
+        8,
+        4,
+        1,
+    ],
+    n_bins=64,
+    name='ZoeDepth',
+    notes='',
+    output_distribution='logbinomial',
+    prefetch=False,
+    pretrained_resource='local::./work_dir/ZoeDepthv1.pt',
+    print_losses=False,
+    project='ZoeDepth',
+    random_crop=False,
+    random_translate=False,
+    root='.',
+    save_dir='',
+    shared_dict='NULL',
+    tags='',
+    train_midas=True,
+    translate_prob=0.2,
+    type='ZoeDepth',
+    uid='NULL',
+    use_amp=False,
+    use_pretrained_midas=True,
+    use_shared_dict=False,
+    validate_every=0.25,
+    version_name='v1',
+    workers=16)