Euler View
1 pending

Run Dossier

Training run for UniFuse

49 metrics · 1,452 train · 0 val

Chart Overlays

Compare this run against its peers

0/59
Add peer runs here to layer their train and val traces onto the current metric charts.
external assumed crashed Apr 12, 21:42 Started 20d 15h ago Python 3.11.7 src/train.py --config configs/model/inverted_joint_metric.yaml --config configs/model/backbone_dinov2_vitl14.yaml --config configs/data/real_drive_sim_euler_loading.yaml --config configs/training/phase_b_joint_structure.yaml --init-from /cluster/scratch/drothenpiele/euler_train/unifuse/checkpoints/geometry-prompt/checkpoint-epoch-2.pt
lr 2e-4batch 4epochs 12precision automodel weather_metric_system wd 1e-4
ID 63160941
unifuse/runs/2026-04-12_21-42-34_395c
Metrics 49/49
depth.train 31
diag 12
consolidation 1
delta_vs_prior
final 3
depth_mae
log_radius_l1
xyz_l1
pred 3
depth_mae
log_radius_l1
xyz_l1
prior 3
depth_mae
log_radius_l1
xyz_l1
structure 2
depth_l1
edge_l1
loss 18
consolidation 3
low_frequency
structure_align
structure_align_edge
final 4
confidence
depth
log_radius
xyz
pred 1
log_radius
prior 4
aux_metric_probe
aux_metric_probe_confidence
confidence
log_radius
structure 5
confidence
depth
depth_denoise_mse
edge
rgb_denoise_mse
no stage 1
total
stat 1
confidence
rgb.train 3
clear
confidence
rgb_denoise_mse
sys.train 15
gpu_mem_total_gb
gpu_mem_used_gb
gpu_mem_util_pct
gpu_util_pct
lr
lr.default
lr.default_no_decay
lr.geometry_encoder_adapter
lr.geometry_encoder_adapter_no_decay
lr.joint_structure_expert_heads
lr.joint_structure_expert_heads_no_decay
lr.joint_structure_expert_model
lr.joint_structure_expert_model_no_decay
lr.prompt_encoder
lr.prompt_encoder_no_decay
49 selected
X-Axis
Y-Scale
Series
Group
Namespace

Delta Vs Prior

↓ kind=diagstage=consolidation
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Depth Mae

↓ meters kind=diagstage=final
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Log Radius L1

↓ kind=diagstage=final
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Xyz L1

↓ meters kind=diagstage=final
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Depth Mae

↓ meters kind=diagstage=pred
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Log Radius L1

↓ kind=diagstage=pred
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Xyz L1

↓ meters kind=diagstage=pred
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Depth Mae

↓ meters kind=diagstage=prior
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Log Radius L1

↓ kind=diagstage=prior
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Xyz L1

↓ meters kind=diagstage=prior
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Depth L1

↓ kind=diagstage=structure
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Edge L1

↓ kind=diagstage=structure
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Low Frequency

↓ kind=lossstage=consolidation
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Structure Align

↓ kind=lossstage=consolidation
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Structure Align Edge

↓ kind=lossstage=consolidation
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Confidence

↑ kind=lossstage=final
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Depth

↓ meters kind=lossstage=final
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Log Radius

↓ kind=lossstage=final
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Xyz

↓ meters kind=lossstage=final
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Log Radius

↓ kind=lossstage=pred
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Aux Metric Probe

kind=lossstage=prior
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Aux Metric Probe Confidence

kind=lossstage=prior
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Confidence

↑ kind=lossstage=prior
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Log Radius

↓ kind=lossstage=prior
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Confidence

↑ kind=lossstage=structure
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Depth

↓ meters kind=lossstage=structure
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Depth Denoise Mse

kind=lossstage=structure
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Edge

↓ kind=lossstage=structure
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Rgb Denoise Mse

kind=lossstage=structure
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Total

↓ kind=loss
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Confidence

↑ kind=statstage=structure
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Clear

kind=lossstage=structure
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Confidence

↑ kind=lossstage=structure
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Rgb Denoise Mse

kind=lossstage=structure
T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Gpu Mem Total Gb

T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Gpu Mem Used Gb

T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Gpu Mem Util Pct

T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Gpu Util Pct

T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Lr

T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Lr.Default

T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Lr.Default No Decay

T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Lr.Geometry Encoder Adapter

T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Lr.Geometry Encoder Adapter No Decay

T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Lr.Joint Structure Expert Heads

T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Lr.Joint Structure Expert Heads No Decay

T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Lr.Joint Structure Expert Model

T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Lr.Joint Structure Expert Model No Decay

T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Lr.Prompt Encoder

T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

Lr.Prompt Encoder No Decay

T
2026-04-12_21-42-34_395c train (smoothed) —
2026-04-12_21-42-34_395c train (raw) —

No output snapshots found for this run.

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

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