r2s-cur-d025-dec8-9a606d4-ef2b28d0-47646604
cancelled —
Hypothesis
With 8 decoder layers, dropout 0.25 transcribes validation renders better than v9. Pass requires the paired 95% bootstrap interval on the per-icon reduction in mean 72 px pixel error on the 339 validation renders to exclude zero against each normal v9 run (RTX 4080 original, H100 rerun, B300 seeds 2 and 3) and, for depth arms, a direction consistent across seeds against dropout 0.2 at 6 layers. Its greedy latency on the RTX 4080 is recorded.
State transitions
- running2026-10-30T18:15:02Z
- cancelled2026-10-30T19:14:28Zcancelled_after_2_min_replaced_by_wave_3b_120k_replications_of_r2s-cur-d02-s120k
Run record
Verbatim from runs/r2s-cur-d025-dec8-9a606d4-ef2b28d0-47646604/run.yaml, the record committed before launch.
schema_version
2
run_id
r2s-cur-d025-dec8-9a606d4-ef2b28d0-47646604
state
cancelled
planned_at
2026-10-30T18:15:02Z
hypothesis
With 8 decoder layers, dropout 0.25 transcribes validation renders better than v9. Pass requires the paired 95% bootstrap interval on the per-icon reduction in mean 72 px pixel error on the 339 validation renders to exclude zero against each normal v9 run (RTX 4080 original, H100 rerun, B300 seeds 2 and 3) and, for depth arms, a direction consistent across seeds against dropout 0.2 at 6 layers. Its greedy latency on the RTX 4080 is recorded.
parent_run
r2s-cur-d02-dec8
git_commit
9a606d47e8a674cddbe38159cae29eb99fd081b0
dirty_patch_sha256
config
configs/render2svg/cur-d025-dec8.yaml
config_sha256
ef2b28d09125d7b1…config_resolved
model
image_size
144
d_model
256
heads
8
encoder_layers
2
decoder_layers
8
feedforward
1024
dropout
0.25
metric
true
fourier
10
order
path
sources
0
training
steps
60000
batch_size
32
learning_rate
0.0005
weight_decay
0.01
warmup_steps
1000
seed
7001
eval_every
2000
eval_icons
64
train_icons
None
evaluate_on_train
false
augment_variants
0
augment_seed
9001
augment_mirror
0.5
augment_max_shift
48
augment_colour
0.1
augment_online
true
compose_probability
0.5
compose_parts
4
augment_original
0.1
loader_workers
16
extra_training
bf16
true
final_eval_icons
339
clip_icons
32
dataset
pilot_config
configs/learning/openmoji-g1-geometric-gate-v16.yaml
cache_sha256
476466042da98d72…train_icons
2681
augmented_variants
0
extra_training
None
augmented_sha256
None
evaluated_on
primary/validation
seed
7001
determinism
seeded; cuDNN and SDPA kernels not forced deterministic
model_parameters
11132706
command
.venv/bin/python -m mojidiff.learning.render2svg --config configs/render2svg/cur-d025-dec8.yaml
outputs
run_dir
runs/r2s-cur-d025-dec8-9a606d4-ef2b28d0-47646604
checkpoints
/home/eai-jdkxz/eai-team11/mojidiff/cache/runs/r2s-cur-d025-dec8-9a606d4-ef2b28d0-47646604
notes
cancelled_at
2026-10-30T19:14:28Z
cancel_reason
cancelled_after_2_min_replaced_by_wave_3b_120k_replications_of_r2s-cur-d02-s120k
slurm_job
13980