omnisvg-n1-image-finetune-greedy-097b037-32icons-9b9b1699
completed —
Measured result
Verbatim from the run's summary.json.
checks
clip_gain_over_control
false
ended_rate
false
reference_top1_rate
false
clip
repo
openai/clip-vit-base-patch32
revision
3d74acf9a28c67741b2f4f2ea7635f0aaf6f0268
clip_to_caption
mean
0.23391288006678224
median
0.2306956872344017
n
32
clip_to_reference
mean
0.8563928715884686
median
0.8664350807666779
n
32
codec_valid_rate
0.25
config_sha256
ffdd05fa66357bb6…control
clip_to_reference_mean
0.8563928715884686
control_clip_to_reference_mean
0.9046997493132949
control_codec_valid_rate
0.140625
control_reference_top1_rate
0.609375
control_rows_sha256
5cabd027e8c4e617…gain_ci95
-0.07607764776330442, -0.01923629671800882
icons_improved
8
mean_gain
-0.048306877724826336
paired_icons
32
report_root
reports/learning/omnisvg-n1-image-control
criteria
clip_gain_over_control_ci_excludes_zero
true
min_ended_rate
0.9
min_reference_top1_rate
0.75
decoded_rate
1.0
drawings
32
ended_rate
0.125
failures
encode:program needs 1 contours and up to 511 segments
2
encode:program needs 1 contours and up to 769 segments
1
encode:program needs 1 contours and up to 845 segments
1
encode:program needs 112 contours and up to 10 segments
1
encode:program needs 126 contours and up to 7 segments
1
encode:program needs 2 contours and up to 500 segments
1
encode:program needs 2 contours and up to 501 segments
2
encode:program needs 3 contours and up to 1006 segments
1
encode:program needs 5 contours and up to 471 segments
1
encode:program needs 5 contours and up to 474 segments
1
encode:program needs 52 contours and up to 11 segments
1
encode:program needs 59 contours and up to 22 segments
1
encode:program needs 6 contours and up to 477 segments
1
encode:program needs 6 contours and up to 918 segments
1
encode:program needs 64 contours and up to 10 segments
1
encode:program needs 76 contours and up to 8 segments
1
encode:program needs 85 contours and up to 12 segments
1
normalize:geometry_error: M has sets of 2 args, 1 invalid
2
pack:program needs 157 packed segments but capacity is 128
1
pack:program needs 571 packed segments but capacity is 128
1
pack:program needs 999 packed segments but capacity is 128
1
icons
32
max_new_tokens
2048
mean_snap_distance_rgb
4.667461976221594
median_paths
4.0
median_tokens
2048.0
model
adapter
data/processed/prior/omnisvg-n1-image-finetune/adapter.zip
adapter_sha256
15b8be9922c3a678…checkpoint_sha256
39f36759fc677b32…fine_tuned
true
parameters
3846914048
repo
OmniSVG/OmniSVG1.1_4B
revision
117d4c4541839e02943554fe71c0a0d54fcef819
predeclared_outcome
falsified
prompt_style
image
reference_rank
chance_top1
0.03125
mean
7.46875
scored
32
top1_rate
0.4375
top5_rate
0.625
rows_sha256
ecb04df39b0980c8…samples_per_icon
1
sampling
greedy
true
repetition_penalty
1.0
schema_version
1
sheet_sha256
1f77101403cc5863…study_version
omnisvg-n1-image-finetune-greedy
timing
generate_seconds
1134.4750232100778
load_seconds
37.50987054902362
peak_vram_gib
8.074842929840088
seconds_per_drawing
35.45234447531493
Visual output
State transitions
- planned2026-09-22T16:19:56Z
- completed2026-09-22T18:48:34Zgreedy without the penalty: 12.5% end, the right icon first for 14 of 32 (0.44), similarity 0.856, paired gain -0.048 with an interval of -0.076 to -0.019; the sheet shows the icon's first parts and then a loop
Run record
This run has no run.yaml. What follows is the identity and configuration carried by its rows in state/runs.jsonl, the append-only registry.
run_id
omnisvg-n1-image-finetune-greedy-097b037-32icons-9b9b1699
config
configs/learning/omnisvg-n1-image-finetune-greedy.yaml
outputs
report_root
reports/learning/omnisvg-n1-image-finetune-greedy