[Spec Decode] Unified Parallel Drafting (#32887)

Signed-off-by: Benjamin Chislett <bchislett@nvidia.com>
This commit is contained in:
Benjamin Chislett
2026-02-05 12:37:18 -05:00
committed by GitHub
parent 5b2a9422f0
commit af3162d3aa
14 changed files with 1085 additions and 392 deletions

View File

@@ -204,7 +204,7 @@ def test_mtp_propose(num_speculative_tokens, monkeypatch):
target_positions=target_positions,
target_hidden_states=target_hidden_states,
next_token_ids=next_token_ids,
last_token_indices=None,
token_indices_to_sample=None,
common_attn_metadata=common_attn_metadata,
sampling_metadata=sampling_metadata,
)