r2s-overfit4-70ac6df-2a313ad8-47646604
failed —
Hypothesis
A 7.6M-parameter render-to-program encoder-decoder, trained on four OpenMoji icons, decodes those four icons greedily from their 144 px renders to programs whose 72 px renders are within 0.01 mean RGB error of the originals, in under 1,000 steps.
Visual output
Written result
Failed while writing its record. record write failed after training and evaluation: yaml could not represent torch.__version__ (TorchVersion); latency and final summary lost. Training metrics preserved in metrics.jsonl: exact reconstruction (pixel error 0.0, free-token accuracy 1.0) of all 4 icons from step 100.
Rerun under a new identity after the fix.
State transitions
- running2026-09-27T19:55:36Z
- failed2026-09-27T19:56:23Zrecord write failed after training and evaluation: yaml could not represent torch.__version__ (TorchVersion); latency and final summary lost. Training metrics preserved in metrics.jsonl: exact reconstruction (pixel error 0.0, free-token accuracy 1.0) of all 4 icons from step 100.
Run record
Verbatim from runs/r2s-overfit4-70ac6df-2a313ad8-47646604/run.yaml, the record committed before launch.
schema_version
2
run_id
r2s-overfit4-70ac6df-2a313ad8-47646604
state
failed
planned_at
2026-09-27T19:55:36Z
hypothesis
A 7.6M-parameter render-to-program encoder-decoder, trained on four OpenMoji icons, decodes those four icons greedily from their 144 px renders to programs whose 72 px renders are within 0.01 mean RGB error of the originals, in under 1,000 steps.
parent_run
None
git_commit
70ac6df535a0903ea96f3057a26720662585ebe2
dirty_patch_sha256
config
configs/render2svg/overfit-4.yaml
config_sha256
2a313ad8e603cb7c…config_resolved
model
image_size
144
d_model
256
heads
8
encoder_layers
2
decoder_layers
6
feedforward
1024
dropout
0.0
training
steps
800
batch_size
4
learning_rate
0.001
weight_decay
0.01
warmup_steps
50
seed
7001
eval_every
100
eval_icons
4
train_icons
4
evaluate_on_train
true
augment_mirror
0.0
bf16
true
final_eval_icons
4
clip_icons
0
dataset
pilot_config
configs/learning/openmoji-g1-geometric-gate-v16.yaml
cache_sha256
476466042da98d72…train_icons
4
evaluated_on
primary/train
seed
7001
determinism
seeded; cuDNN and SDPA kernels not forced deterministic
model_parameters
8921634
command
.venv/bin/python -m mojidiff.learning.render2svg --config configs/render2svg/overfit-4.yaml
outputs
run_dir
runs/r2s-overfit4-70ac6df-2a313ad8-47646604
checkpoints
/home/dev/.cache/mojidiff/runs/r2s-overfit4-70ac6df-2a313ad8-47646604
notes
failed_at
2026-09-27T19:56:23Z
failure_reason
record write failed after training and evaluation: yaml could not represent torch.__version__ (TorchVersion); latency and final summary lost. Training metrics preserved in metrics.jsonl: exact reconstruction (pixel error 0.0, free-token accuracy 1.0) of all 4 icons from step 100.