- 29 Aug, 2025 6 commits
-
-
zhuwenwen authored
fix bug in zero-overhead core See merge request dcutoolkit/deeplearing/vllm!192
-
lizhigong authored
-
zhuwenwen authored
-
zhuwenwen authored
-
yangql authored
-
zhuwenwen authored
[fix]优化 block_tables 的生成逻辑,增加对混合情况的检测,确保在存在空和非空块时正确计算最大块长度。 See merge request dcutoolkit/deeplearing/vllm!191
-
- 28 Aug, 2025 1 commit
-
-
laibao authored
-
- 27 Aug, 2025 4 commits
- 26 Aug, 2025 6 commits
- 25 Aug, 2025 3 commits
- 22 Aug, 2025 2 commits
- 21 Aug, 2025 2 commits
- 20 Aug, 2025 6 commits
- 19 Aug, 2025 3 commits
- 18 Aug, 2025 7 commits
-
-
zhuwenwen authored
fix issue from merge See merge request dcutoolkit/deeplearing/vllm!184
-
zhuwenwen authored
[fix]解决v1 deepseek cudagraph模式显存占用增长 See merge request dcutoolkit/deeplearing/vllm!183
-
lizhigong authored
-
王敏 authored
-
zhuwenwen authored
-
zhuwenwen authored
-
zhuwenwen authored
[fix]修复mtp eager模式下显存占用增加问题 See merge request dcutoolkit/deeplearing/vllm!180
-