vllm/v1/worker/gpu_input_batch.py · 19b927e52df8400084df1c8116af7d6f0a5f5d15 · OpenDAS / vllm_cscc · GitLab

Find file Blame History Permalink

[Core] Use individual MM items in P0/P1 cache and model runner (#22570) · 19b927e5
Cyrus Leung authored Aug 13, 2025
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
19b927e5

gpu_input_batch.py 33 KB