CI Remove bnb docker image build from GH workflow (#2859)

In #2858, the bitsandbytes tests and Dockerfiles were removed from the
PEFT repo. However, I forgot to update the GitHub workflow that builds
these images. This PR now fixes that.

Also changed the image used by the torch.compile workflow, as it was
using a bnb image.

Author: BenjaminBossan

.github/workflows/build_docker_images.yml

@@ -81,70 +81,3 @@ jobs:
           title: 🤗 Results of the PEFT-GPU docker build
           status: ${{ job.status }}
           slack_token: ${{ secrets.SLACK_CIFEEDBACK_BOT_TOKEN }}
-
-  latest-cuda-bnb-source:
-    name: "Latest Peft GPU + bnb source [dev]"
-    runs-on:
-      group: aws-general-8-plus
-    steps:
-      - name: Set up Docker Buildx
-        uses: docker/setup-buildx-action@b5ca514318bd6ebac0fb2aedd5d36ec1b5c232a2  # v3.10.0
-      - name: Check out code
-        uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8  # v5.0.0
-        with:
-          persist-credentials: false
-      - name: Login to DockerHub
-        uses: docker/login-action@74a5d142397b4f367a81961eba4e8cd7edddf772  # v3.4.0
-        with:
-          username: ${{ secrets.DOCKERHUB_USERNAME }}
-          password: ${{ secrets.DOCKERHUB_PASSWORD }}
-
-      - name: Build and Push GPU
-        uses: docker/build-push-action@14487ce63c7a62a4a324b0bfb37086795e31c6c1  # v6.16.0
-        with:
-          context: ./docker/peft-gpu-bnb-source
-          push: true
-          tags: huggingface/peft-gpu-bnb-source
-
-      - name: Post to Slack
-        if: always()
-        uses: huggingface/hf-workflows/.github/actions/post-slack@3f88d63d3761558a32e8e46fc2a8536e04bb2aea  # main from Feb 2025-02-24
-        with:
-          slack_channel: ${{ env.CI_SLACK_CHANNEL }}
-          title: 🤗 Results of the PEFT-GPU (bnb source / HF latest) docker build
-          status: ${{ job.status }}
-          slack_token: ${{ secrets.SLACK_CIFEEDBACK_BOT_TOKEN }}
-
-  latest-cuda-bnb-source-latest:
-    name: "Latest Peft GPU + bnb source [accelerate / peft / transformers latest]"
-    runs-on:
-      group: aws-general-8-plus
-    steps:
-      - name: Set up Docker Buildx
-        uses: docker/setup-buildx-action@b5ca514318bd6ebac0fb2aedd5d36ec1b5c232a2  # v3.10.0
-      - name: Check out code
-        uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8  # v5.0.0
-        with:
-          persist-credentials: false
-      - name: Login to DockerHub
-        uses: docker/login-action@74a5d142397b4f367a81961eba4e8cd7edddf772  # v3.4.0
-        with:
-          username: ${{ secrets.DOCKERHUB_USERNAME }}
-          password: ${{ secrets.DOCKERHUB_PASSWORD }}
-
-      - name: Build and Push GPU
-        uses: docker/build-push-action@14487ce63c7a62a4a324b0bfb37086795e31c6c1  # v6.16.0
-        with:
-          context: ./docker/peft-gpu-bnb-latest
-          push: true
-          tags: huggingface/peft-gpu-bnb-latest
-
-      - name: Post to Slack
-        if: always()
-        uses: huggingface/hf-workflows/.github/actions/post-slack@3f88d63d3761558a32e8e46fc2a8536e04bb2aea  # main from Feb 2025-02-24
-        with:
-          slack_channel: ${{ env.CI_SLACK_CHANNEL }}
-          title: 🤗 Results of the PEFT-GPU (bnb source / HF source) docker build
-          status: ${{ job.status }}
-          slack_token: ${{ secrets.SLACK_CIFEEDBACK_BOT_TOKEN }}
-

.github/workflows/torch_compile_tests.yml

@@ -26,10 +26,10 @@ jobs:
     env:
       PEFT_DEBUG_WITH_TORCH_COMPILE: 1
       CUDA_VISIBLE_DEVICES: "0"
-      TEST_TYPE: "single_gpu_huggingface/peft-gpu-bnb-latest:latest"
+      TEST_TYPE: "single_gpu_huggingface/peft-gpu:latest"
       USE_PYTORCH_NIGHTLY: "${{ github.event.inputs.pytorch_nightly }}"
     container:
-      image: "huggingface/peft-gpu-bnb-latest:latest"
+      image: "huggingface/peft-gpu:latest"
       options: --gpus all --shm-size "16gb" --ipc host -v /mnt/cache/.cache/huggingface:/mnt/cache/
     defaults:
       run: