Make ResBlock's conditioning-injection mechanism selectable (gitea #34) #49

Merged

lars merged 1 commits from fix/issue-34 into master 2026-08-14 09:52:19 +02:00

Conversation 0 Commits 1 Files Changed 8

+280 -14

lars commented 2026-08-14 09:46:58 +02:00

Owner

Copy Link

Copy Source

ResBlock injected conditioning exactly one way — h = linear1(h) +
cond_proj(cond), a conditional bias, the weakest standard option for a
model whose entire job is to be conditional. Adds BLOCK_REGISTRY
(giant/model/layers.py), mirroring the TRUNK_REGISTRY/ROUTER_REGISTRY
registry+factory idiom (gitea #33), with two new drop-in alternatives:
FilmResBlock (per-channel scale+shift modulating the norm output,
zero-init so conditioning has no effect at construction) and
AdaLNResBlock (DiT-style AdaLN-Zero — the norm's own affine is replaced
by a conditioning-derived scale/shift, plus a zero-init gate on the
residual branch, making the block the exact identity function at init).

Selected per stage via a new stage{1,2}_model.trunk.block_conditioning
config leaf ("add" | "film" | "adaln", default "add"), threaded through
build_trunk/build_expert_body/RoutedTrunk and the three stage model
constructors. Default stays "add" and ResBlock's body is unchanged, so
existing configs/checkpoints are bit-identical to before this change.

Decided during planning: the new field lives on the existing TrunkConfig
rather than a new top-level block/blocks config section; the WGAN
CriticModel (which builds its own ResBlock stack outside TRUNK_REGISTRY)
and the issue's mentioned blocks.norm/blocks.activation axes are both
left out of scope.

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

ResBlock injected conditioning exactly one way — h = linear1(h) + cond_proj(cond), a conditional bias, the weakest standard option for a model whose entire job is to be conditional. Adds BLOCK_REGISTRY (giant/model/layers.py), mirroring the TRUNK_REGISTRY/ROUTER_REGISTRY registry+factory idiom (gitea #33), with two new drop-in alternatives: FilmResBlock (per-channel scale+shift modulating the norm output, zero-init so conditioning has no effect at construction) and AdaLNResBlock (DiT-style AdaLN-Zero — the norm's own affine is replaced by a conditioning-derived scale/shift, plus a zero-init gate on the residual branch, making the block the exact identity function at init). Selected per stage via a new stage{1,2}_model.trunk.block_conditioning config leaf ("add" | "film" | "adaln", default "add"), threaded through build_trunk/build_expert_body/RoutedTrunk and the three stage model constructors. Default stays "add" and ResBlock's body is unchanged, so existing configs/checkpoints are bit-identical to before this change. Decided during planning: the new field lives on the existing TrunkConfig rather than a new top-level block/blocks config section; the WGAN CriticModel (which builds its own ResBlock stack outside TRUNK_REGISTRY) and the issue's mentioned blocks.norm/blocks.activation axes are both left out of scope. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

lars added 1 commit 2026-08-14 09:46:59 +02:00

Make ResBlock's conditioning-injection mechanism selectable (gitea #34) ... 0f95e0eaae

ResBlock injected conditioning exactly one way — h = linear1(h) +
cond_proj(cond), a conditional bias, the weakest standard option for a
model whose entire job is to be conditional. Adds BLOCK_REGISTRY
(giant/model/layers.py), mirroring the TRUNK_REGISTRY/ROUTER_REGISTRY
registry+factory idiom (gitea #33), with two new drop-in alternatives:
FilmResBlock (per-channel scale+shift modulating the norm output,
zero-init so conditioning has no effect at construction) and
AdaLNResBlock (DiT-style AdaLN-Zero — the norm's own affine is replaced
by a conditioning-derived scale/shift, plus a zero-init gate on the
residual branch, making the block the exact identity function at init).

Selected per stage via a new stage{1,2}_model.trunk.block_conditioning
config leaf ("add" | "film" | "adaln", default "add"), threaded through
build_trunk/build_expert_body/RoutedTrunk and the three stage model
constructors. Default stays "add" and ResBlock's body is unchanged, so
existing configs/checkpoints are bit-identical to before this change.

Decided during planning: the new field lives on the existing TrunkConfig
rather than a new top-level block/blocks config section; the WGAN
CriticModel (which builds its own ResBlock stack outside TRUNK_REGISTRY)
and the issue's mentioned blocks.norm/blocks.activation axes are both
left out of scope.

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

lars merged commit c8f52259d6 into master 2026-08-14 09:52:19 +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#49