tzhouamandClaude Fable 5.1 3085f50160 [Edge][CPU] Compile the per-sub-step lm head + top-k Gumbel sampling + embedding + projection into one graph ...

Noise is drawn eagerly with the request generators (same RNG consumption as the
eager loop) and passed in, so the compiled post-step is generator-free and
shape-static; used on the fp32/torchao CPU path outside stored (top-p) mode.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>

2026-09-10 03:12:40 +08:00

..

model_loader

[Ming] Cleanup Ming-family shared modules (#6119)

2026-08-19 14:13:40 +08:00

models

[Edge][CPU] Compile the per-sub-step lm head + top-k Gumbel sampling + embedding + projection into one graph

2026-09-10 03:12:40 +08:00

quantization

[Quantization][Qwen3-Omni] Enable NVFP4 W4A4 serving on Blackwell (#4025)

2026-06-18 10:01:46 +08:00

stage_input_processors

[Model] Preserve MiniCPM-o Talker full attention until capacity (#6799)

2026-09-06 22:25:35 +08:00

conftest.py

[Test] Unify L2/L3 test layout, Buildkite steps, and test helpers (#2556)

2026-05-08 14:12:23 +08:00

helpers.py

[Test] Unify L2/L3 test layout, Buildkite steps, and test helpers (#2556)

2026-05-08 14:12:23 +08:00

test_cfg_pairing.py

[Model] Add MiniMax Music 3 text-to-music (#6186)

2026-08-18 04:33:13 +00:00