YiYi Xu 5dc153bc4a support loading pipeline from transformer style (flat) repo (#14096) ...

* Support root-hosted weights + add _allow_files download hook

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

2026-07-02 08:04:31 -10:00

..

fixtures

[chore] update to 2026 finally. (#14079)

2026-06-29 20:20:44 -04:00

hooks

[tests] migrate test_hooks.py to pytest (#13242)

2026-06-05 10:33:45 +05:30

lora

[Tests] Skip layerwise casting tests on devices without float8_e4m3fn support (#14073)

2026-06-29 06:38:00 -07:00

models

[Tests] Skip layerwise casting tests on devices without float8_e4m3fn support (#14073)

2026-06-29 06:38:00 -07:00

modular_pipelines

[chore] update to 2026 finally. (#14079)

2026-06-29 20:20:44 -04:00

others

[chore] update to 2026 finally. (#14079)

2026-06-29 20:20:44 -04:00

pipelines

support loading pipeline from transformer style (flat) repo (#14096)

2026-07-02 08:04:31 -10:00

quantization

[chore] update to 2026 finally. (#14079)

2026-06-29 20:20:44 -04:00

remote

Sunset Python 3.8 & get rid of explicit typing exports where possible (#12524)

2026-02-13 18:16:51 +05:30

schedulers

[discrete diffusion] Add DiffusionGemma pipeline and schedulers (#13986)

2026-06-25 08:22:16 +02:00

single_file

Use device_map="auto" in single file tests to support large models on limited GPU memory (#13816)

2026-06-11 09:09:13 +05:30

__init__.py

fix issues with loading, add test for pipeline

2022-06-07 15:40:36 +02:00

conftest.py

[chore] update to 2026 finally. (#14079)

2026-06-29 20:20:44 -04:00

testing_utils.py

Fix model cuda tests (#13975)

2026-06-23 17:42:36 +05:30