Fix cu_num_generated_tokens slicing logic in LogprobsLists.slice() method (#28214)

Signed-off-by: Bradley <bradley.b.pitt@gmail.com>

tests/v1/test_outputs.py

+# SPDX-License-Identifier: Apache-2.0
+# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
+from unittest import TestCase
+
+from vllm.v1.outputs import LogprobsLists
+
+
+class TestLogprobsLists(TestCase):
+    def setUp(self):
+        self.logprobsLists = LogprobsLists(
+            logprob_token_ids=[
+                [1, 2],  # Request 0 token 0
+                [3, 4],  # Request 0 token 1
+                [5, 6],  # Request 1 token 0
+                [7, 8],  # Request 1 token 1
+                [9, 10],  # Request 1 token 2
+                [11, 12],  # Request 2 token 0
+                [13, 14],  # Request 2 token 1
+                [15, 16],  # Request 2 token 2
+                [17, 18],  # Request 2 token 3
+            ],
+            logprobs=[
+                [0.1, 0.2],
+                [0.3, 0.4],
+                [0.5, 0.6],
+                [0.7, 0.8],
+                [0.9, 1.0],
+                [1.1, 1.2],
+                [1.3, 1.4],
+                [1.5, 1.6],
+                [1.7, 1.8],
+            ],
+            sampled_token_ranks=[1, 3, 5, 7, 9, 11, 13, 15, 17],
+            cu_num_generated_tokens=[0, 2, 5, 9],
+        )
+
+    def test_slice_without_cu_num_generated_tokens(self):
+        """Test slicing without cu_num_generated_tokens"""
+        logprobsLists = LogprobsLists(
+            logprob_token_ids=[[1], [2], [3]],
+            logprobs=[[0.1], [0.2], [0.3]],
+            sampled_token_ranks=[1, 2, 3],
+            cu_num_generated_tokens=None,
+        )
+
+        sliced = logprobsLists.slice(1, 3)
+        assert sliced.logprob_token_ids == [[2], [3]]
+        assert sliced.logprobs == [[0.2], [0.3]]
+        assert sliced.sampled_token_ranks == [2, 3]
+        assert sliced.cu_num_generated_tokens is None
+
+    def test_slice_from_start(self):
+        """Test slicing from the start position"""
+        sliced = self.logprobsLists.slice(0, 2)
+        assert len(sliced.logprob_token_ids) == 5
+        assert sliced.logprob_token_ids == [
+            [1, 2],
+            [3, 4],
+            [5, 6],
+            [7, 8],
+            [9, 10],
+        ]
+        assert sliced.cu_num_generated_tokens == [0, 2, 5]
+
+    def test_slice_from_middle(self):
+        """Test slicing from the middle position"""
+        sliced = self.logprobsLists.slice(1, 3)
+        assert len(sliced.logprob_token_ids) == 7
+        assert sliced.logprob_token_ids == [
+            [5, 6],
+            [7, 8],
+            [9, 10],
+            [11, 12],
+            [13, 14],
+            [15, 16],
+            [17, 18],
+        ]
+        assert sliced.cu_num_generated_tokens == [0, 3, 7]
+
+    def test_slice_single_request(self):
+        """Test slicing a single request"""
+        sliced = self.logprobsLists.slice(1, 2)
+        assert len(sliced.logprob_token_ids) == 3
+        assert sliced.logprob_token_ids == [[5, 6], [7, 8], [9, 10]]
+        assert sliced.cu_num_generated_tokens == [0, 3]
+
+    def test_slice_last_request(self):
+        """Test slicing the last request"""
+        sliced = self.logprobsLists.slice(2, 3)
+        assert len(sliced.logprob_token_ids) == 4
+        assert sliced.logprob_token_ids == [[11, 12], [13, 14], [15, 16], [17, 18]]
+        assert sliced.cu_num_generated_tokens == [0, 4]
+
+    def test_slice_all_requests(self):
+        """Test slicing all requests (full slice)"""
+        sliced = self.logprobsLists.slice(0, 3)
+        assert len(sliced.logprob_token_ids) == 9  # All tokens
+        assert sliced.logprob_token_ids == self.logprobsLists.logprob_token_ids
+        assert (
+            sliced.cu_num_generated_tokens == self.logprobsLists.cu_num_generated_tokens
+        )

vllm/v1/outputs.py

@@ -30,16 +30,23 @@ class LogprobsLists(NamedTuple):
         if self.cu_num_generated_tokens:
             start = self.cu_num_generated_tokens[start_req_idx]
             end = self.cu_num_generated_tokens[end_req_idx]
+            # Recompute cumulative array starting from 0
+            cu_num_offset = self.cu_num_generated_tokens[start_req_idx]
+            sliced_cu_num_generated_tokens = [
+                cu_num - cu_num_offset
+                for cu_num in self.cu_num_generated_tokens[
+                    start_req_idx : end_req_idx + 1
+                ]
+            ]
         else:
             start = start_req_idx
             end = end_req_idx
+            sliced_cu_num_generated_tokens = None
         return LogprobsLists(
             self.logprob_token_ids[start:end],
             self.logprobs[start:end],
             self.sampled_token_ranks[start:end],
-            self.cu_num_generated_tokens[start_req_idx:end_req_idx]
-            if self.cu_num_generated_tokens
-            else None,
+            sliced_cu_num_generated_tokens,
         )