Repository navigation

[PyTorch] Support split Linear parameters with torch.compile - #3623

Open

pggPL wants to merge 30 commits into

NVIDIA:mainfrom

pggPL:split_parameters_linear

Open

[PyTorch] Support split Linear parameters with torch.compile#3623
pggPL wants to merge 30 commits into
NVIDIA:mainfrom
pggPL:split_parameters_linear

Commits on Oct 5, 2026
- [PyTorch] Avoid concatenation copies for compiled split Linear parameters
  pggPL
  committed
  fdaea9e
- [PyTorch] Make deferred parameter concatenation explicit
  pggPL
  committed
  289cff1
- [PyTorch] Simplify split Linear parameter handling
  pggPL
  committed
  43a33bb
- [PyTorch] Limit the split-parameter API refactor
  pggPL
  committed
  5d5c180
Commits on Oct 12, 2026
- Merge main into split_parameters_linear
  pggPL
  committed
  bdb5bc1
- [PyTorch] Keep the frozen-weight bias fix out of split parameter support
  pggPL
  committed
  73d58af
Commits on Oct 14, 2026
- [PyTorch] Read split parameter shape and gradient flags directly
  pggPL
  committed
  122a2c9
- [PyTorch] Pass split parameters as tensor lists
  pggPL
  committed
  156576e
- [PyTorch] Use representation-neutral names for Linear operands
  pggPL
  committed
  dd01b3a
Commits on Oct 16, 2026
- [PyTorch] Pass split Linear weights in separate tensor-list fields
  pggPL
  committed
  88b47c3
- [PyTorch] Skip empty split-parameter bookkeeping in eager Linear
  pggPL
  committed
  bcf13f6
- [PyTorch] Make deferred Linear parameter lists optional
  pggPL
  committed
  623baf8
- [PyTorch] Trim split parameter tests
  pggPL
  committed
  7b28093
Commits on Oct 21, 2026
- [PyTorch] Materialize split Linear parameters inside eager forward
  pggPL
  committed
  40974d3
Commits on Oct 27, 2026
- Merge upstream main into split_parameters_linear
  pggPL
  committed
  ddb3eaa
- Merge upstream main and preserve split Linear documentation
  pggPL
  committed
  f5a13d2
Commits on Oct 28, 2026
- Merge upstream main into split_parameters_linear
  pggPL
  committed
  acfce62
- [PyTorch] Materialize split LayerNormLinear parameters in eager forward
  pggPL
  committed
  303b093
- [PyTorch] Consolidate split parameter integration tests
  pggPL
  committed
  61e1854
- [PyTorch] Avoid repeated split-parameter validation
  pggPL
  committed
  576b505
- Simplify deferred split parameter selection
  pggPL
  committed
  06415ad
- Clarify split parameter preparation names
  pggPL
  committed
  d661987
- Detect split backward inputs from saved metadata
  pggPL
  committed
  15b9973
- Reuse adapter metadata to restore split parameters
  pggPL
  committed
  63ccf35
- Cover split parameters in existing module tests
  pggPL
  committed
  4fa64a8
- Remove added split LayerNormLinear sanity scenarios
  pggPL
  committed
  10105b3
- Keep module output and gradient checks local to compile tests
  pggPL
  committed
  b04c1e3
Commits on Oct 29, 2026

Commits

Commits on Oct 5, 2026

Commits on Oct 12, 2026

Commits on Oct 14, 2026

Commits on Oct 16, 2026

Commits on Oct 21, 2026

Commits on Oct 27, 2026

Commits on Oct 28, 2026

Commits on Oct 29, 2026