LLM inference in C/C++
C++ ggml
ci : use t4-medium for cuda jobs (#29842)
Sigbjørn Skjæret4ebdf2c74acc
diff --git a/.github/workflows/ci-self-hosted-cuda.yml b/.github/workflows/ci-self-hosted-cuda.ymlindex 5951b52d37..24f830a9f0 100644--- a/.github/workflows/ci-self-hosted-cuda.yml+++ b/.github/workflows/ci-self-hosted-cuda.yml@@ -45,7 +45,7 @@ env:jobs:gpu-cuda:- runs-on: "hf-jobs-t4-small:cuda13"+ runs-on: "hf-jobs-t4-medium:cuda13"steps:- name: Clonediff --git a/.github/workflows/server-self-hosted.yml b/.github/workflows/server-self-hosted.ymlindex 8dc4637c45..614fd32f5b 100644--- a/.github/workflows/server-self-hosted.yml+++ b/.github/workflows/server-self-hosted.yml@@ -102,7 +102,7 @@ jobs:PYTEST_WORKERS=1 ./tests.shserver-cuda:- runs-on: "hf-jobs-t4-small:cuda13"+ runs-on: "hf-jobs-t4-medium:cuda13"steps:- name: Clone