[MRV2][BugFix] Fix KV connector handling in spec decode case (#43719)
Signed-off-by: Nick Hill <nickhill123@gmail.com> Co-authored-by: Wentao Ye <44945378+yewentao256@users.noreply.github.com>
This commit is contained in:
@@ -168,7 +168,7 @@ def test_v2_sample_tokens_runs_eplb_on_non_last_pp_rank(monkeypatch):
|
||||
slot_mappings_by_layer=None,
|
||||
hidden_states=None,
|
||||
aux_hidden_states=None,
|
||||
kv_connector_output=None,
|
||||
finished_req_ids=set(),
|
||||
num_tokens_across_dp=None,
|
||||
)
|
||||
runner.postprocess = lambda *args, **kwargs: events.append("postprocess")
|
||||
|
||||
Reference in New Issue
Block a user