trigger ci on A10

Author: ydshieh

.github/workflows/model_jobs.yml

@@ -107,9 +107,9 @@ jobs:
         run: |
           echo "${{ inputs.machine_type }}"
 
-          if [ "${{ inputs.machine_type }}" = "aws-g4dn-4xlarge-cache" ]; then
+          if [ "${{ inputs.machine_type }}" = "aws-g5-4xlarge-cache" ]; then
             machine_type=single-gpu
-          elif [ "${{ inputs.machine_type }}" = "aws-g4dn-12xlarge-cache" ]; then
+          elif [ "${{ inputs.machine_type }}" = "aws-g5-12xlarge-cache" ]; then
             machine_type=multi-gpu
           else
             machine_type=${{ inputs.machine_type }}

.github/workflows/self-scheduled-caller.yml

@@ -7,7 +7,7 @@ on:
     - cron: "17 2 * * *"
   push:
     branches:
-      - run_scheduled_ci*
+      - trigger_ci_on_a10
   workflow_dispatch:
     inputs:
       prev_workflow_run_id:
@@ -25,7 +25,7 @@ on:
 # Used for `push` to easily modiffy the target workflow runs to compare against
 env:
     prev_workflow_run_id: ""
-    other_workflow_run_id: ""
+    other_workflow_run_id: "15770139098"
 
 
 jobs:
@@ -50,70 +50,9 @@ jobs:
     uses: ./.github/workflows/self-scheduled.yml
     with:
       job: run_models_gpu
-      slack_report_channel: "#transformers-ci-daily-models"
+      slack_report_channel: "#transformers-ci-dummy"
       runner: daily-ci
       docker: huggingface/transformers-all-latest-gpu
       ci_event: Daily CI
       report_repo_id: hf-internal-testing/transformers_daily_ci
     secrets: inherit
-
-  torch-pipeline:
-    name: Torch pipeline CI
-    uses: ./.github/workflows/self-scheduled.yml
-    with:
-      job: run_pipelines_torch_gpu
-      slack_report_channel: "#transformers-ci-daily-pipeline-torch"
-      runner: daily-ci
-      docker: huggingface/transformers-pytorch-gpu
-      ci_event: Daily CI
-      report_repo_id: hf-internal-testing/transformers_daily_ci
-    secrets: inherit
-
-  example-ci:
-    name: Example CI
-    uses: ./.github/workflows/self-scheduled.yml
-    with:
-      job: run_examples_gpu
-      slack_report_channel: "#transformers-ci-daily-examples"
-      runner: daily-ci
-      docker: huggingface/transformers-all-latest-gpu
-      ci_event: Daily CI
-      report_repo_id: hf-internal-testing/transformers_daily_ci
-    secrets: inherit
-
-  trainer-fsdp-ci:
-    name: Trainer/FSDP CI
-    uses: ./.github/workflows/self-scheduled.yml
-    with:
-      job: run_trainer_and_fsdp_gpu
-      slack_report_channel: "#transformers-ci-daily-training"
-      runner: daily-ci
-      docker: huggingface/transformers-all-latest-gpu
-      ci_event: Daily CI
-      report_repo_id: hf-internal-testing/transformers_daily_ci
-    secrets: inherit
-
-  deepspeed-ci:
-    name: DeepSpeed CI
-    uses: ./.github/workflows/self-scheduled.yml
-    with:
-      job: run_torch_cuda_extensions_gpu
-      slack_report_channel: "#transformers-ci-daily-training"
-      runner: daily-ci
-      docker: huggingface/transformers-pytorch-deepspeed-latest-gpu
-      ci_event: Daily CI
-      working-directory-prefix: /workspace
-      report_repo_id: hf-internal-testing/transformers_daily_ci
-    secrets: inherit
-
-  quantization-ci:
-    name: Quantization CI
-    uses: ./.github/workflows/self-scheduled.yml
-    with:
-      job: run_quantization_torch_gpu
-      slack_report_channel: "#transformers-ci-daily-quantization"
-      runner: daily-ci
-      docker: huggingface/transformers-quantization-latest-gpu
-      ci_event: Daily CI
-      report_repo_id: hf-internal-testing/transformers_daily_ci
-    secrets: inherit

.github/workflows/self-scheduled.yml

@@ -53,7 +53,7 @@ jobs:
     name: Setup
     strategy:
       matrix:
-        machine_type: [aws-g4dn-4xlarge-cache, aws-g4dn-12xlarge-cache]
+        machine_type: [aws-g5-4xlarge-cache, aws-g5-12xlarge-cache]
     runs-on:
       group: '${{ matrix.machine_type }}'
     container:
@@ -111,7 +111,7 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        machine_type: [aws-g4dn-4xlarge-cache, aws-g4dn-12xlarge-cache]
+        machine_type: [aws-g5-4xlarge-cache, aws-g5-12xlarge-cache]
         slice_id: ${{ fromJSON(needs.setup.outputs.slice_ids) }}
     uses: ./.github/workflows/model_jobs.yml
     with:
@@ -129,7 +129,7 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        machine_type: [aws-g4dn-4xlarge-cache, aws-g4dn-12xlarge-cache]
+        machine_type: [aws-g5-4xlarge-cache, aws-g5-12xlarge-cache]
         slice_id: [0, 1]
     uses: ./.github/workflows/model_jobs.yml
     with:
@@ -147,7 +147,7 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        machine_type: [aws-g4dn-4xlarge-cache, aws-g4dn-12xlarge-cache]
+        machine_type: [aws-g5-4xlarge-cache, aws-g5-12xlarge-cache]
     runs-on:
       group: '${{ matrix.machine_type }}'
     container:
@@ -181,9 +181,9 @@ jobs:
         run: |
           echo "${{ matrix.machine_type }}"
 
-          if [ "${{ matrix.machine_type }}" = "aws-g4dn-4xlarge-cache" ]; then
+          if [ "${{ matrix.machine_type }}" = "aws-g5-4xlarge-cache" ]; then
             machine_type=single-gpu
-          elif [ "${{ matrix.machine_type }}" = "aws-g4dn-12xlarge-cache" ]; then
+          elif [ "${{ matrix.machine_type }}" = "aws-g5-12xlarge-cache" ]; then
             machine_type=multi-gpu
           else
             machine_type=${{ matrix.machine_type }}
@@ -215,7 +215,7 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        machine_type: [aws-g4dn-4xlarge-cache]
+        machine_type: [aws-g5-4xlarge-cache]
     runs-on:
       group: '${{ matrix.machine_type }}'
     container:
@@ -249,9 +249,9 @@ jobs:
         run: |
           echo "${{ matrix.machine_type }}"
 
-          if [ "${{ matrix.machine_type }}" = "aws-g4dn-4xlarge-cache" ]; then
+          if [ "${{ matrix.machine_type }}" = "aws-g5-4xlarge-cache" ]; then
             machine_type=single-gpu
-          elif [ "${{ matrix.machine_type }}" = "aws-g4dn-12xlarge-cache" ]; then
+          elif [ "${{ matrix.machine_type }}" = "aws-g5-12xlarge-cache" ]; then
             machine_type=multi-gpu
           else
             machine_type=${{ matrix.machine_type }}
@@ -284,7 +284,7 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        machine_type: [aws-g4dn-4xlarge-cache, aws-g4dn-12xlarge-cache]
+        machine_type: [aws-g5-4xlarge-cache, aws-g5-12xlarge-cache]
     runs-on:
       group: '${{ matrix.machine_type }}'
     container:
@@ -346,9 +346,9 @@ jobs:
         run: |
           echo "${{ matrix.machine_type }}"
 
-          if [ "${{ matrix.machine_type }}" = "aws-g4dn-4xlarge-cache" ]; then
+          if [ "${{ matrix.machine_type }}" = "aws-g5-4xlarge-cache" ]; then
             machine_type=single-gpu
-          elif [ "${{ matrix.machine_type }}" = "aws-g4dn-12xlarge-cache" ]; then
+          elif [ "${{ matrix.machine_type }}" = "aws-g5-12xlarge-cache" ]; then
             machine_type=multi-gpu
           else
             machine_type=${{ matrix.machine_type }}
@@ -383,7 +383,7 @@ jobs:
       fail-fast: false
       matrix:
         folders: ${{ fromJson(needs.setup.outputs.quantization_matrix) }}
-        machine_type: [aws-g4dn-4xlarge-cache, aws-g4dn-12xlarge-cache]
+        machine_type: [aws-g5-4xlarge-cache, aws-g5-12xlarge-cache]
     runs-on:
       group: '${{ matrix.machine_type }}'
     container:
@@ -426,9 +426,9 @@ jobs:
         run: |
           echo "${{ matrix.machine_type }}"
 
-          if [ "${{ matrix.machine_type }}" = "aws-g4dn-4xlarge-cache" ]; then
+          if [ "${{ matrix.machine_type }}" = "aws-g5-4xlarge-cache" ]; then
             machine_type=single-gpu
-          elif [ "${{ matrix.machine_type }}" = "aws-g4dn-12xlarge-cache" ]; then
+          elif [ "${{ matrix.machine_type }}" = "aws-g5-12xlarge-cache" ]; then
             machine_type=multi-gpu
           else
             machine_type=${{ matrix.machine_type }}

utils/split_model_tests.py

@@ -62,4 +62,5 @@
         start = end
         end = start + num_jobs_per_splits + (1 if idx < num_jobs % args.num_splits else 0)
         model_splits.append(d[start:end])
+    model_splits = [["models/vit"], ["models/clip"]]
     print(model_splits)