[Core] Refactoring sampler and support prompt logprob for chunked prefill (#4309)

This commit is contained in:
SangBin Cho
2024-04-26 22:02:02 +09:00
committed by GitHub
parent a88081bf76
commit 603ad84815
18 changed files with 859 additions and 630 deletions

View File

@@ -219,7 +219,7 @@ class _AsyncLLMEngine(LLMEngine):
request_outputs = self._process_model_outputs(
output, scheduler_outputs.scheduled_seq_groups,
scheduler_outputs.ignored_seq_groups)
scheduler_outputs.ignored_seq_groups, seq_group_metadata_list)
# Log stats.
if self.log_stats: