No jobs

Run Dossier

Training run for SD 2.1. Mari

22 metrics · 40,693 train · 84 val

Chart Overlays

Compare this run against its peers

0/63
Add peer runs here to layer their train and val traces onto the current metric charts.
external assumed crashed May 15, 22:26 Duration: 5s Python 3.11.7 /cluster/home/drothenpiele/.cache/venv/mari/bin/euler-eval /cluster/home/drothenpiele/models/stable_diffusion_mari/mari/.euler-launches/euler_launch_86bc2b2a-0d6a-44e3-85aa-1b07a811c566/config.json --mask-sky --rgb-fid-backend clean-fid --benchmark-depth-range 0.01 80.0 --batch-size 16 -v
lr 4e-5batch 16epochs 200precision fp16model stable-diffusion-2-1 seed 42wd 0grad accum 1warmup 0.05000ema nograd clip 1
ID 66667314
mari/runs/2026-05-15_22-26-22_1533
Metrics 22/22
depth.train 5
diag 3
depth_abs_rel
depth_delta1
depth_mae
loss 2
depth
depth_head
rgb.train 11
diag 2
image_psnr
image_ssim
loss 7
dehaze
preserve
total
visibility
visibility_aux
visibility_rank
visibility_tv
stat 2
visibility_mean
visibility_std
sys.train 6
gpu_mem_total_gb
gpu_mem_used_gb
gpu_mem_util_pct
gpu_util_pct
lr
visibility_aux_scale
X-Axis
Y-Scale
Series
Group
Namespace

Depth Abs Rel

kind=diag
V
2026-05-15_22-26-22_1533 val (smoothed)
2026-05-15_22-26-22_1533 val (raw)

Depth Delta1

kind=diag
V
2026-05-15_22-26-22_1533 val (smoothed)
2026-05-15_22-26-22_1533 val (raw)

Depth Mae

kind=diag
V
2026-05-15_22-26-22_1533 val (smoothed)
2026-05-15_22-26-22_1533 val (raw)

Depth

kind=loss
T
2026-05-15_22-26-22_1533 train (smoothed)
2026-05-15_22-26-22_1533 train (raw)

Depth Head

kind=loss
T
2026-05-15_22-26-22_1533 train (smoothed)
2026-05-15_22-26-22_1533 train (raw)

Image Psnr

dB kind=diag
V
2026-05-15_22-26-22_1533 val (smoothed)
2026-05-15_22-26-22_1533 val (raw)

Image Ssim

kind=diag
V
2026-05-15_22-26-22_1533 val (smoothed)
2026-05-15_22-26-22_1533 val (raw)

Dehaze

kind=loss
T
2026-05-15_22-26-22_1533 train (smoothed)
2026-05-15_22-26-22_1533 train (raw)

Preserve

kind=loss
T
2026-05-15_22-26-22_1533 train (smoothed)
2026-05-15_22-26-22_1533 train (raw)

Total

kind=loss
T
2026-05-15_22-26-22_1533 train (smoothed)
2026-05-15_22-26-22_1533 train (raw)

Visibility

kind=loss
T
2026-05-15_22-26-22_1533 train (smoothed)
2026-05-15_22-26-22_1533 train (raw)

Visibility Aux

kind=loss
T
2026-05-15_22-26-22_1533 train (smoothed)
2026-05-15_22-26-22_1533 train (raw)

Visibility Rank

kind=loss
T
2026-05-15_22-26-22_1533 train (smoothed)
2026-05-15_22-26-22_1533 train (raw)

Visibility Tv

kind=loss
T
2026-05-15_22-26-22_1533 train (smoothed)
2026-05-15_22-26-22_1533 train (raw)

Visibility Mean

kind=stat
T V
2026-05-15_22-26-22_1533 train (smoothed)
2026-05-15_22-26-22_1533 val (smoothed)
2026-05-15_22-26-22_1533 train (raw)
2026-05-15_22-26-22_1533 val (raw)

Visibility Std

kind=stat
T V
2026-05-15_22-26-22_1533 train (smoothed)
2026-05-15_22-26-22_1533 val (smoothed)
2026-05-15_22-26-22_1533 train (raw)
2026-05-15_22-26-22_1533 val (raw)

Gpu Mem Total Gb

T
2026-05-15_22-26-22_1533 train (smoothed)
2026-05-15_22-26-22_1533 train (raw)

Gpu Mem Used Gb

T
2026-05-15_22-26-22_1533 train (smoothed)
2026-05-15_22-26-22_1533 train (raw)

Gpu Mem Util Pct

T
2026-05-15_22-26-22_1533 train (smoothed)
2026-05-15_22-26-22_1533 train (raw)

Gpu Util Pct

T
2026-05-15_22-26-22_1533 train (smoothed)
2026-05-15_22-26-22_1533 train (raw)

Lr

T
2026-05-15_22-26-22_1533 train (smoothed)
2026-05-15_22-26-22_1533 train (raw)

Visibility Aux Scale

T
2026-05-15_22-26-22_1533 train (smoothed)
2026-05-15_22-26-22_1533 train (raw)

No output snapshots found for this run.

Outputs are generated during training and saved to outputs/epoch_N_step_M/ directories.

epoch 1 / step 2442 Checkpoint #3399
5/16/2026, 9:55:11 AM
/cluster/scratch/drothenpiele/euler_train/mari/checkpoints/joint-dehaze-metric-2026-05-15-22-26-22-1533/checkpoint-1
epoch 3 / step 4884 Checkpoint #3400
5/16/2026, 9:55:11 AM
/cluster/scratch/drothenpiele/euler_train/mari/checkpoints/joint-dehaze-metric-2026-05-15-22-26-22-1533/checkpoint-3
epoch 5 / step 7326 Checkpoint #3401
5/16/2026, 9:55:11 AM
/cluster/scratch/drothenpiele/euler_train/mari/checkpoints/joint-dehaze-metric-2026-05-15-22-26-22-1533/checkpoint-5
epoch 7 / step 9768 Checkpoint #3402
5/16/2026, 9:55:11 AM
/cluster/scratch/drothenpiele/euler_train/mari/checkpoints/joint-dehaze-metric-2026-05-15-22-26-22-1533/checkpoint-7
epoch 9 / step 12210 Checkpoint #3403
5/16/2026, 9:55:11 AM
/cluster/scratch/drothenpiele/euler_train/mari/checkpoints/joint-dehaze-metric-2026-05-15-22-26-22-1533/checkpoint-9
epoch 11 / step 14652 Checkpoint #3404
5/16/2026, 9:55:11 AM
/cluster/scratch/drothenpiele/euler_train/mari/checkpoints/joint-dehaze-metric-2026-05-15-22-26-22-1533/checkpoint-11
epoch 13 / step 17094 Checkpoint #3417
5/16/2026, 11:42:33 AM
/cluster/scratch/drothenpiele/euler_train/mari/checkpoints/joint-dehaze-metric-2026-05-15-22-26-22-1533/checkpoint-13
epoch 15 / step 19536 Checkpoint #3432
5/16/2026, 11:16:43 PM
/cluster/scratch/drothenpiele/euler_train/mari/checkpoints/joint-dehaze-metric-2026-05-15-22-26-22-1533/checkpoint-15
epoch 17 / step 21978 Checkpoint #3433
5/16/2026, 11:16:43 PM
/cluster/scratch/drothenpiele/euler_train/mari/checkpoints/joint-dehaze-metric-2026-05-15-22-26-22-1533/checkpoint-17
epoch 19 / step 24420 Checkpoint #3434
5/16/2026, 11:16:43 PM
/cluster/scratch/drothenpiele/euler_train/mari/checkpoints/joint-dehaze-metric-2026-05-15-22-26-22-1533/checkpoint-19
epoch 21 / step 26862 Checkpoint #3435
5/16/2026, 11:16:43 PM
/cluster/scratch/drothenpiele/euler_train/mari/checkpoints/joint-dehaze-metric-2026-05-15-22-26-22-1533/checkpoint-21
epoch 23 / step 29304 Checkpoint #3436
5/16/2026, 11:16:43 PM
/cluster/scratch/drothenpiele/euler_train/mari/checkpoints/joint-dehaze-metric-2026-05-15-22-26-22-1533/checkpoint-23
Producer launch exports are available. Manage launch-owned exports here for quick reference.
Open Launch

Inherited Launch Exports

These exports are published by the run's producer launch.

Published

The producer launch does not publish any exports yet.

Run-Owned Exports

Publish direct filesystem paths here.

Published

No run-owned exports are published yet.

Raw Artifacts

Run-owned exports are typically direct paths, so there are no captured artifacts to publish from here.

Euler View - ML Experiment Monitor