Add configs/baseline.toml as the kept reference model #58

Merged

lars merged 1 commits from add/baseline-config into master 2026-08-14 17:37:52 +02:00

Conversation 0 Commits 1 Files Changed 1

+129

lars commented 2026-08-14 17:37:24 +02:00

Owner

Copy Link

Copy Source

A fixed comparison point for future architecture variants, so each
experimental axis (routed trunk, WGAN generators, attention history,
shared conditioning) is a single edit away from one known config.

flow/flow autoregressive, hidden_dim 512 / 6 blocks per stage, physical
conditioning, no router, 7.70M params. Chosen by ranking the five runs in
analysis_runs/ by mean Jensen-Shannon divergence against the Geant4
reference: unrouted flow wins (0.172) over routed flow (0.197/0.200) and
both WGAN runs (0.218/0.234), with the lead concentrated in per-event
total deposited energy and the per-PDG marginals.

batch_size 36864 is sized for one L40S on deepthought2 from a measured
linear fit of this config's training step (reserved MiB = 0.9736 * bs +
115), giving ~36 GiB, 78% of the card.

The comments record two measured facts that are easy to get wrong:
WGAN is slower to train than flow (n_critic plus the gradient-penalty
double-backward), its advantage being inference-only; and
sample_secondaries_ar loops over all k_max slots unconditionally rather
than short-circuiting on n_sec, which is what makes the autoregressive
decoder the dominant cost on both axes.

Co-Authored-By: Claude Opus 5 noreply@anthropic.com

A fixed comparison point for future architecture variants, so each experimental axis (routed trunk, WGAN generators, attention history, shared conditioning) is a single edit away from one known config. flow/flow autoregressive, hidden_dim 512 / 6 blocks per stage, physical conditioning, no router, 7.70M params. Chosen by ranking the five runs in analysis_runs/ by mean Jensen-Shannon divergence against the Geant4 reference: unrouted flow wins (0.172) over routed flow (0.197/0.200) and both WGAN runs (0.218/0.234), with the lead concentrated in per-event total deposited energy and the per-PDG marginals. batch_size 36864 is sized for one L40S on deepthought2 from a measured linear fit of this config's training step (reserved MiB = 0.9736 * bs + 115), giving ~36 GiB, 78% of the card. The comments record two measured facts that are easy to get wrong: WGAN is slower to *train* than flow (n_critic plus the gradient-penalty double-backward), its advantage being inference-only; and sample_secondaries_ar loops over all k_max slots unconditionally rather than short-circuiting on n_sec, which is what makes the autoregressive decoder the dominant cost on both axes. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

lars added 1 commit 2026-08-14 17:37:26 +02:00

Add configs/baseline.toml as the kept reference model ... d858226294

A fixed comparison point for future architecture variants, so each
experimental axis (routed trunk, WGAN generators, attention history,
shared conditioning) is a single edit away from one known config.

flow/flow autoregressive, hidden_dim 512 / 6 blocks per stage, physical
conditioning, no router, 7.70M params. Chosen by ranking the five runs in
analysis_runs/ by mean Jensen-Shannon divergence against the Geant4
reference: unrouted flow wins (0.172) over routed flow (0.197/0.200) and
both WGAN runs (0.218/0.234), with the lead concentrated in per-event
total deposited energy and the per-PDG marginals.

batch_size 36864 is sized for one L40S on deepthought2 from a measured
linear fit of this config's training step (reserved MiB = 0.9736 * bs +
115), giving ~36 GiB, 78% of the card.

The comments record two measured facts that are easy to get wrong:
WGAN is slower to *train* than flow (n_critic plus the gradient-penalty
double-backward), its advantage being inference-only; and
sample_secondaries_ar loops over all k_max slots unconditionally rather
than short-circuiting on n_sec, which is what makes the autoregressive
decoder the dominant cost on both axes.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

lars merged commit a1ecf0df1d into master 2026-08-14 17:37:52 +02:00

lars deleted branch add/baseline-config 2026-08-14 17:38:00 +02:00

Sign in to join this conversation.

Reviewers

No Reviewers

Labels

Clear labels

architecture

bug

Silent/wrong behavior, no crash

experiment

feature

New feature matching the v0.3 redesign

investigation

modularity

Making the network genuinely pluggable

performance

phase2

validation

No Label

Milestone

No items

No Milestone

Projects

Clear projects

No project

Assignees

Clear assignees

lars (Lars Bogner)

No Assignees

1 Participants

Notifications

Subscribe

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: lars/giant#58