[Split 2/N of #3430] feat(mHC): GPT model wiring - #4945

Draft

Connor-XY wants to merge 20 commits into

NVIDIA:mainfrom

Connor-XY:mhc-pr2-gpt-wiring

Commits on Aug 13, 2026

[Split 1/N of #3430] feat(mHC): basic pytorch implementation of manifold hyper connection

committed
512cf77
test(mHC): make test_transformer_layer self-contained for split #1
Connor-XY
committed
c7b88dd
test(mHC): update test_hybrid_moe_model GOLDEN_CONFIG for new mHC fields
Connor-XY
committed
1e3cea5
test(mHC): add direct coverage for BDA recompute path and block-level forward
Connor-XY
committed
fd2a61d
style: collapse short attention_mask to one line for black
Connor-XY
committed
c3586ab
refactor(mHC): address mathemakitten review on PR #4531

Connor-XY
and
claude
committed
fcbc630
refactor(mHC): drop dead is_last_layer_in_recompute_block kwarg pop

Connor-XY
and
claude
committed
970bcde
refactor(mHC): extract attention hooks to dedup _forward_attention

Connor-XY
and
claude
committed
62af205
refactor(mHC): thread h_res/h_post via attn_state slot, not instance attrs

Connor-XY
and
claude
committed
ffd0f60
style(mHC): restore _set_proj_residual context comment

Connor-XY
and
claude
committed
f6383ec
refactor(mHC): extract MLP hooks to dedup _forward_mlp

Connor-XY
and
claude
committed
a4a68a7
style(mHC): shorten docstring line to fit 100-char pylint limit

Connor-XY
and
claude
committed
0ae94c8
fix(mHC): only thread mhc_recompute_manager when manager exists

Connor-XY
and
claude
committed
5a7c634
refactor(mHC): add back-compat shim for renamed _forward_post_mlp hook

Connor-XY
and
claude
committed
787ee11
test(mHC): pass pipeline rank to offload manager
Connor-XY
committed
673c771
fix(mhc): update CUDA graph module API

FDecaYed
authored and
Connor-XY
committed
5e253ae
refactor(mHC): address review feedback
Connor-XY
committed
5dbf710
fix(mHC): drop a bad rebase hunk and guard unsupported mHC configs
Connor-XY
committed
bdfed9a
[Split 2/N of #3430] feat(mHC): GPT model wiring

committed
6bbe5d7
refactor(mHC): rename spec/config plumbing to the mhc_* vocabulary
Connor-XY
committed
af41e41

Navigation Menu

Uh oh!

[Split 2/N of #3430] feat(mHC): GPT model wiring - #4945

[Split 2/N of #3430] feat(mHC): GPT model wiring#4945
Connor-XY wants to merge 20 commits into
NVIDIA:mainfrom
Connor-XY:mhc-pr2-gpt-wiring

Commits on Aug 13, 2026

[Split 1/N of #3430] feat(mHC): basic pytorch implementation of manifold hyper connection

test(mHC): make test_transformer_layer self-contained for split #1

test(mHC): update test_hybrid_moe_model GOLDEN_CONFIG for new mHC fields

test(mHC): add direct coverage for BDA recompute path and block-level forward

style: collapse short attention_mask to one line for black

refactor(mHC): address mathemakitten review on PR #4531

refactor(mHC): drop dead is_last_layer_in_recompute_block kwarg pop

refactor(mHC): extract attention hooks to dedup _forward_attention

refactor(mHC): thread h_res/h_post via attn_state slot, not instance attrs

style(mHC): restore _set_proj_residual context comment

refactor(mHC): extract MLP hooks to dedup _forward_mlp

style(mHC): shorten docstring line to fit 100-char pylint limit

fix(mHC): only thread mhc_recompute_manager when manager exists

refactor(mHC): add back-compat shim for renamed _forward_post_mlp hook

test(mHC): pass pipeline rank to offload manager

fix(mhc): update CUDA graph module API

refactor(mHC): address review feedback

fix(mHC): drop a bad rebase hunk and guard unsupported mHC configs

[Split 2/N of #3430] feat(mHC): GPT model wiring

refactor(mHC): rename spec/config plumbing to the mhc_* vocabulary