vllm/v1/worker/gpu_model_runner.py · c42ff4f4fdc4a4d48ccef18b8067995f6c19e6ec · OpenDAS / vllm_cscc · GitLab

Find file Blame History Permalink

[BugFix][torch.compile] KV scale calculation issues with FP8 quantization (#25513) · c42ff4f4
Adrian Abeyta authored Sep 29, 2025
```
Signed-off-by: adabeyta <aabeyta@redhat.com>
```
c42ff4f4

gpu_model_runner.py 193 KB