1 pending

Run Dossier

Training run for MERGE

8 metrics · 21,136 train · 0 val

Chart Overlays

Compare this run against its peers

0/12
Add peer runs here to layer their train and val traces onto the current metric charts.
external assumed crashed Feb 21, 21:08 Started 70d 15h ago Python 3.9.18 merge/train_dehaze_depth.py --pretrained_model_name_or_path=PixArt-alpha/PixArt-XL-2-512x512 --config=/cluster/home/drothenpiele/models/MERGE/.euler-launches/euler_launch_eb8a8b75-fca4-427e-b823-3f9fe2b578dc/train_dehaze_depth.yaml --output_dir=/cluster/scratch/drothenpiele/MERGE/static_prompt_unfrozen_2_300m_rds_8 --train_batch_size=1 --gradient_accumulation_steps=16 --num_train_epochs=1000 --learning_rate=0.0002 --lr_scheduler=linear --lr_warmup_steps=50 --mixed_precision=fp16 --seed=2024 --num_layers=14 --num_attention_heads=16 --attention_head_dim=72 --ff_mult=1 --converter_init_type=pretrained --multi_res_noise=True --lambda_dehaze=1 --lambda_depth=1 --checkpointing_steps=1000 --validation_steps=100000 --report_to=wandb --wandb_name=dehaze_depth_vkitti --tracker_project_name=merge_dehaze_depth_rds --base_data_dir=/cluster/scratch/drothenpiele/data/real-drive-sim/sample_10_zips --dataloader_num_workers=2 --max_grad_norm=1 --num_unfrozen_backbone_layers=2 --embeddings_dir=/cluster/scratch/drothenpiele/MERGE/embeddings/text --run_name=Unfrozen-2-300m --runlog_dir=/cluster/work/igp_psr/drothenpiele/out/train/MERGE
lr 2e-4precision fp16 seed 2024grad accum 16grad clip 1
ID 58074258 Job MERGE_d_d Part gpuhe.120h Node eu-g6-077 CPU 8 GPU nvidia_geforce_rtx_4090:1
train/MERGE/runs/2026-02-21_21-08-57_8666
Metrics 8/8
gpu 4
mem 3
total/gb
used/gb
util/pct
util/pct
dehaze/loss
depth/loss
lr
total/loss
X-Axis
Y-Scale
Series

Dehaze Loss

T
2026-02-21_21-08-57_8666 train (smoothed)
2026-02-21_21-08-57_8666 train (raw)

Depth Loss

T
2026-02-21_21-08-57_8666 train (smoothed)
2026-02-21_21-08-57_8666 train (raw)

Gpu Mem Total Gb

T
2026-02-21_21-08-57_8666 train (smoothed)
2026-02-21_21-08-57_8666 train (raw)

Gpu Mem Used Gb

T
2026-02-21_21-08-57_8666 train (smoothed)
2026-02-21_21-08-57_8666 train (raw)

Gpu Mem Util Pct

T
2026-02-21_21-08-57_8666 train (smoothed)
2026-02-21_21-08-57_8666 train (raw)

Gpu Util Pct

T
2026-02-21_21-08-57_8666 train (smoothed)
2026-02-21_21-08-57_8666 train (raw)

Lr

T
2026-02-21_21-08-57_8666 train (smoothed)
2026-02-21_21-08-57_8666 train (raw)

Total Loss

T
2026-02-21_21-08-57_8666 train (smoothed)
2026-02-21_21-08-57_8666 train (raw)

No output snapshots found for this run.

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

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