Peiyuan Zhang 82e00615e5 [feat]: wire v2 VideoGenerator to real torch backend

2026-07-04 19:31:55 +00:00

..

__init__.py

[feat] multi-backend dispatch substrate (device/arch/kernel registries)

2026-07-04 17:15:34 +00:00

accel.py

[refactor]: remove training semantics from v2 inference contracts

2026-07-04 19:31:35 +00:00

cpu.py

[refactor] v2: group flat top-level into planes; isolate vendored under _vendor/

2026-07-04 17:15:34 +00:00

GPU_BRINGUP.md

[fix] correct GPU adapters against real fastvideo API (cross-check findings)

2026-07-04 17:15:34 +00:00

torch_backend.py

[feat]: wire v2 VideoGenerator to real torch backend

2026-07-04 19:31:55 +00:00

torch_cuda.py

[misc] v2: simplify docstrings/comments + drop deleted-design-doc citations

2026-07-04 17:15:34 +00:00

torch_kernels.py

[misc] v2: full pre-commit clean (ruff UP038/SIM/UP031 + mypy annotations)

2026-07-04 17:15:34 +00:00

toy.py

[refactor]: remove training semantics from v2 inference contracts

2026-07-04 19:31:35 +00:00