Add files via upload

config/cfg_kitti_autoencoder.py

@@ -0,0 +1,66 @@
+DEPTH_LAYERS = 50
+POSE_LAYERS = 18
+FRAME_IDS = [0]
+IMGS_PER_GPU = 5
+HEIGHT = 256
+WIDTH = 800
+
+data = dict(
+    name = 'kitti',
+    split = 'exp',
+    height = HEIGHT,
+    width = WIDTH,
+    frame_ids = FRAME_IDS,
+    in_path = '/node01_data5/kitti_raw',
+    gt_depth_path = '/node01_data5/monodepth2-test/monodepth2/gt_depths.npz',
+    png = False,
+    stereo_scale = False,
+)
+
+model = dict(
+    name = 'autoencoder',
+    depth_num_layers = DEPTH_LAYERS,
+    pose_num_layers = POSE_LAYERS,
+    frame_ids = FRAME_IDS,
+    imgs_per_gpu = IMGS_PER_GPU,
+    height = HEIGHT,
+    width = WIDTH,
+    scales = [0, 1, 2, 3],
+    min_depth = 0.1,
+    max_depth = 100.0,
+    depth_pretrained_path = '/node01/jobs/io/pretrained/checkpoints/resnet/resnet{}.pth'.format(DEPTH_LAYERS),
+    pose_pretrained_path =  '/node01/jobs/io/pretrained/checkpoints/resnet/resnet{}.pth'.format(POSE_LAYERS),
+    automask = True,
+    disp_norm = True,
+    use_min_construct = True,
+    dis=0.001,
+    cvt=0.001,
+)
+
+# resume_from = '/node01_data5/monodepth2-test/model/ms/ms.pth'
+resume_from = None
+finetune = None
+total_epochs = 30
+imgs_per_gpu = IMGS_PER_GPU
+learning_rate = 1e-4
+workers_per_gpu = 4
+validate = False
+
+optimizer = dict(type='Adam', lr=learning_rate, weight_decay=0)
+optimizer_config = dict(grad_clip=dict(max_norm=35, norm_type=2))
+lr_config = dict(
+    policy='step',
+    warmup='linear',
+    warmup_iters=500,
+    warmup_ratio=1.0 / 3,
+    step=[10,20],
+    gamma=0.5,
+)
+
+checkpoint_config = dict(interval=1)
+log_config = dict(interval=50,
+                  hooks=[dict(type='TextLoggerHook'),])
+dist_params = dict(backend='nccl')
+log_level = 'INFO'
+load_from = None
+workflow = [('train', 1)]