Giancarlo Delfin c32e97602d [Model Runner V2] Enable forcing a specific acceptance rate during rejection sampling (#38045) ...

Signed-off-by: Giancarlo Delfin <gdelfin@inferact.ai>

2026-03-26 13:38:12 -07:00

..

__init__.py

…

test_acceptance_length.py

[Hardware] Replace torch.cuda.device_count/current_device/set_device API (#36145)

2026-03-12 07:57:47 -07:00

test_eagle_step_kernel.py

feat(spec_decode): fuse EAGLE step slot mapping and metadata updates (#33503)

2026-03-11 04:35:33 +00:00

test_eagle.py

[Async][Spec Decoding] Zero-bubble async scheduling + spec decoding (#32951)

2026-03-23 15:37:22 -04:00

test_extract_hidden_states.py

[Async][Spec Decoding] Zero-bubble async scheduling + spec decoding (#32951)

2026-03-23 15:37:22 -04:00

test_max_len.py

…

test_mtp.py

[BugFix] Add support for MTP num_speculative_tokens > 1 with sparse MLA (#34552)

2026-03-03 07:21:57 -08:00

test_ngram.py

…

test_speculators_eagle3.py

…

test_synthetic_rejection_sampler_utils.py

[Model Runner V2] Enable forcing a specific acceptance rate during rejection sampling (#38045)

2026-03-26 13:38:12 -07:00

test_tree_attention.py

[ROCm][CI] Fix spec decode logprobs flakiness and parametrize tree attention backends (#34599)

2026-02-20 20:25:23 -08:00