No jobs
Back to SD 2.1. Mari

Train: Metric Depth V6 6f2e9dce

Supersedes Train: Metric Depth V5. Makes the scalar depth head the primary metric-depth path, separates valid-depth mask thresholds from log-normalization bounds, and starts V6 with conservative fusion/visibility defaults.

training failed

Output

pending 6f2e9dce
Launch Output
Waiting for SLURM job to start...

Failure Metadata

failed
interactive.monitor.poll INT_EXIT_CODE_NONZERO

Interactive command exited with code 255

Context JSON
{
  "reason": null,
  "exitCode": 255,
  "remotePid": 2015117,
  "computePid": null,
  "allocationJobId": 2694947
}

Datasets

5
Dataset clear_root_rds
Type rgb
Split full
Path /cluster/work/igp_psr/drothenpiele/data/rds/rgb.zip/
Dataset depth_root_rds
Type depth
Split full
Path /cluster/work/igp_psr/drothenpiele/data/rds/depth.zip/
Dataset train_hazy_root_rds
Type rgb
Split train
Path /cluster/work/igp_psr/drothenpiele/data/rds/gloomy/noise_model/foggy_rgb.zip/
Dataset val_hazy_root_rds
Type rgb
Split val
Path /cluster/work/igp_psr/drothenpiele/data/rds/gloomy/noise_model/foggy_rgb.zip/
Dataset clear_root_vkitti2
Dataset muses
Type rgb
Split fog_day
Path /cluster/work/igp_psr/drothenpiele/data/muses/frame_camera_trainvaltest.zip/

Execution Artifacts

2
run.sh

Published Exports

Launch-Owned Exports

These semantic handles are what downstream pipelines should resolve against. Auto-published exports come from config-template metadata captured on this launch.

Published

This launch does not publish any exports yet.

Raw Artifacts

Parameters

167

Typed Parameters

Dataset clear_root_rds
real-drive-sim / full (4)
Dataset depth_root_rds
real-drive-sim / full (5)
Dataset train_hazy_root_rds
real-drive-sim / train (103)
Dataset val_hazy_root_rds
real-drive-sim / val (104)
Dataset clear_root_vkitti2
muses / fog_day (77)
Dataset depth_root_vkitti2
muses / fog_day (77)
Dataset train_hazy_vkitti
muses / fog_day (77)
Dataset val_hazy_vkitti
muses / fog_day (77)

Simple Parameters

clip
true
crop
center
seed
42
tracker
wandb
use_ema
false
gpu_type
rtx_4090:1
job_name
sd_train_concat
norm_max
1
norm_min
-1
run_name
joint-dehaze-metric
tmp_size
10G
adam_8bit
true
data_kind
real_drive_sim
ema_decay
0.9999
ema_dtype
float32
log_every
10
max_depth
800
min_depth
1
adam_beta1
0.9
adam_beta2
0.999
batch_size
16
clear_root
4
depth_mode
metric_log
depth_root
5
image_size
[384, 768]
log_images
true
num_epochs
200
output_dir
/cluster/scratch/drothenpiele/SD21/exp_1
time_limit
2-00:00:00
mem_per_cpu
8G
num_workers
4
adam_epsilon
1e-8
adam_foreach
false
aspect_ratio
[1, 2]
conditioning
concat
lambda_depth
1
min_lr_ratio
0.01
project_name
joint-dehazing
warmup_ratio
0.05
weight_decay
0
cpus_per_task
8
lambda_dehaze
1
learning_rate
0.00004
max_depth_rds
500
max_grad_norm
1
min_depth_rds
0.00001
val_hazy_root
104
freeze_encoder
false
log_visibility
true
num_log_images
4
use_depth_head
true
val_batch_size
10
enable_xformers
true
euler_train_dir
/cluster/work/igp_psr/drothenpiele/data/out/train/sd21-joint
lambda_depth_tv
0
mixed_precision
fp16
prediction_type
v_prediction
source_kind_rds
real_drive_sim
train_hazy_root
103
visibility_cmap
viridis
depth_noise_type
annealed_multires
lr_schedule_type
constant_with_warmup
min_max_quantile
0.02
no_decay_enabled
false
pretrained_model
sd2-community/stable-diffusion-2-1
sky_depth_meters
800
conv_in_init_mode
marigold
joint_weight_psnr
1
joint_weight_ssim
10
lambda_depth_head
0.1
lambda_visibility
0.05
max_depth_vkitti2
300
min_depth_vkitti2
0.00001
sampling_strategy
source_balanced
source_weight_rds
1
depth_ensemble_tol
0.001
val_every_n_epochs
2
depth_ensemble_size
3
num_inference_steps
25
sampling_epoch_size
balanced
save_every_n_epochs
2
source_kind_vkitti2
vkitti2
use_visibility_head
true
warmup_start_factor
0.001
lambda_vae_depth_rec
0.0
lambda_visibility_tv
0.005
sky_depth_meters_rds
800
depth_gradient_robust
charbonnier
depth_multires_levels
4
encoder_learning_rate
0.000015
lambda_depth_gradient
0
source_weight_vkitti2
1
use_cross_task_fusion
false
vae_decoder_trainable
false
visibility_rank_pairs
128
weight_decay_backbone
0
weight_decay_no_decay
0
zero_grad_set_to_none
true
depth_ensemble_max_res
1024
depth_smoothness_space
auto
freeze_cross_attention
true
gradient_checkpointing
true
inference_depth_output
depth_head
lambda_visibility_rank
0.02
visibility_rank_margin
0.05
depth_ensemble_max_iter
2
depth_multires_strength
0.9
depth_smoothness_scales
4
joint_weight_delta1_pct
0.5
keep_last_n_checkpoints
3
validation_depth_output
depth_head
visibility_target_gamma
4
visibility_warmup_steps
1000
cross_task_fusion_blocks
[0]
cross_task_fusion_detach
true
depth_ensemble_reduction
median
depth_normalization_type
log_depth
joint_weight_abs_rel_pct
0.5
sky_depth_meters_vkitti2
800
depth_diagnostics_enabled
true
num_inference_steps_depth
10
vae_decoder_learning_rate
0.000001
visibility_apply_to_depth
false
depth_head_hidden_channels
64
depth_resize_interpolation
bilinear
depth_smoothness_normalize
true
enable_efficient_attention
true
lambda_visibility_preserve
0.05
visibility_apply_to_dehaze
true
visibility_apply_to_fusion
false
visibility_hidden_channels
32
visibility_target_quantile
0.9
weight_decay_depth_decoder
0
checkpoint_selection_metric
"joint" # "joint" or single metric: psnr/ssim/delta1/abs_rel/val/...
depth_decoder_learning_rate
0.00004
gradient_accumulation_steps
1
weight_decay_dehaze_decoder
0
cross_task_fusion_directions
["rgb_to_depth"]
dehaze_decoder_learning_rate
0.00004
depth_smoothness_edge_source
hazy
depth_smoothness_edge_weight
10
lambda_depth_edge_smoothness
0
depth_diagnostics_edge_source
dehazed
depth_diagnostics_edge_weight
10
visibility_decoder_grad_scale
1.0
checkpoint_selection_direction
"auto" # used for single-metric selection: auto|max|min
cross_task_fusion_learned_gate
true
depth_multires_downscale_factor
2
lambda_depth_gt_edge_smoothness
0
cross_task_fusion_gate_init_bias
-4
cross_task_fusion_sender_lowpass
false
cross_task_fusion_hidden_channels
64
depth_diagnostics_normalize_depth
true
lambda_depth_multiscale_smoothness
0
depth_ensemble_regularizer_strength
0.02
depth_diagnostics_road_edge_threshold
0.08
depth_diagnostics_road_bottom_fraction
0.4
cross_task_fusion_sender_lowpass_kernel
3
depth_diagnostics_high_frequency_kernel
5
Raw JSON
{
  "clip": "true",
  "crop": "center",
  "seed": 42,
  "tracker": "wandb",
  "use_ema": "false",
  "gpu_type": "rtx_4090:1",
  "job_name": "sd_train_concat",
  "norm_max": 1,
  "norm_min": -1,
  "run_name": "joint-dehaze-metric",
  "tmp_si...

Events

Launch Events

0
No launch events recorded.
Euler View - ML Experiment Monitor