4.46.3 base

.circleci/parse_test_outputs.py

@@ -67,4 +67,4 @@ def main():
 
 
 if __name__ == "__main__":
-    main()
+    main()

.github/ISSUE_TEMPLATE/bug-report.yml

@@ -1,6 +1,17 @@
 name: "\U0001F41B Bug Report"
 description: Submit a bug report to help us improve transformers
+labels: [ "bug" ]
 body:
+  - type: markdown
+    attributes:
+      value: |
+        Thanks for taking the time to fill out this bug report! 🤗
+
+        Before you submit your bug report:
+
+          - If it is your first time submitting, be sure to check our [bug report guidelines](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#did-you-find-a-bug)
+          - Try our [docs bot](https://huggingface.co/spaces/huggingchat/hf-docs-chat) -- it might be able to help you with your issue
+
   - type: textarea
     id: system-info
     attributes:
@@ -25,26 +36,26 @@ body:
 
         Models:
 
-          - text models: @ArthurZucker 
-          - vision models: @amyeroberts
-          - speech models: @sanchit-gandhi
+          - text models: @ArthurZucker
+          - vision models: @amyeroberts, @qubvel
+          - speech models: @ylacombe, @eustlb
           - graph models: @clefourrier
 
         Library:
 
           - flax: @sanchit-gandhi
           - generate: @zucchini-nlp (visual-language models) or @gante (all others)
-          - pipelines: @Narsil
+          - pipelines: @Rocketknight1
           - tensorflow: @gante and @Rocketknight1
-          - tokenizers: @ArthurZucker
+          - tokenizers: @ArthurZucker and @itazap
           - trainer: @muellerzr @SunMarc
-        
+
         Integrations:
-        
+
           - deepspeed: HF Trainer/Accelerate: @muellerzr
           - ray/raytune: @richardliaw, @amogkam
           - Big Model Inference: @SunMarc
-          - quantization (bitsandbytes, autogpt): @SunMarc
+          - quantization (bitsandbytes, autogpt): @SunMarc @MekkCyber
 
         Documentation: @stevhliu
 

.github/ISSUE_TEMPLATE/i18n.md

@@ -34,7 +34,7 @@ Some notes:
 
 ## Tutorial section
 - [ ] [pipeline_tutorial.md](https://github.com/huggingface/transformers/blob/main/docs/source/en/pipeline_tutorial.md)
-- [ ]  [autoclass_tutorial.md](https://github.com/huggingface/transformers/blob/master/docs/source/autoclass_tutorial.md)
+- [ ]  [autoclass_tutorial.md](https://github.com/huggingface/transformers/blob/main/docs/source/en/autoclass_tutorial.md)
 - [ ]  [preprocessing.md](https://github.com/huggingface/transformers/blob/main/docs/source/en/preprocessing.md)
 - [ ]  [training.md](https://github.com/huggingface/transformers/blob/main/docs/source/en/training.md)
 - [ ]  [accelerate.md](https://github.com/huggingface/transformers/blob/main/docs/source/en/accelerate.md)

.github/PULL_REQUEST_TEMPLATE.md

@@ -40,27 +40,28 @@ members/contributors who may be interested in your PR.
 Models:
 
 - text models: @ArthurZucker
-- vision models: @amyeroberts
-- speech models: @sanchit-gandhi
+- vision models: @amyeroberts, @qubvel
+- speech models: @ylacombe, @eustlb
 - graph models: @clefourrier
 
 Library:
 
 - flax: @sanchit-gandhi
 - generate: @zucchini-nlp (visual-language models) or @gante (all others)
-- pipelines: @Narsil
+- pipelines: @Rocketknight1
 - tensorflow: @gante and @Rocketknight1
 - tokenizers: @ArthurZucker
 - trainer: @muellerzr and @SunMarc
+- chat templates: @Rocketknight1
 
 Integrations:
 
 - deepspeed: HF Trainer/Accelerate: @muellerzr
 - ray/raytune: @richardliaw, @amogkam
 - Big Model Inference: @SunMarc
-- quantization (bitsandbytes, autogpt): @SunMarc 
+- quantization (bitsandbytes, autogpt): @SunMarc @MekkCyber
 
-Documentation: @stevhliu and @MKhalusova
+Documentation: @stevhliu
 
 HF projects:
 

.github/workflows/TROUBLESHOOT.md

@@ -0,0 +1,9 @@
+# Troubleshooting
+
+This is a document explaining how to deal with various issues on github-actions self-hosted CI. The entries may include actual solutions or pointers to Issues that cover those.
+
+## GitHub Actions (self-hosted CI)
+
+* Deepspeed
+
+  - if jit build hangs, clear out `rm -rf ~/.cache/torch_extensions/` reference: https://github.com/huggingface/transformers/pull/12723

.github/workflows/add-model-like.yml

@@ -0,0 +1,80 @@
+name: Add model like runner
+
+on:
+  push:
+    branches:
+      - none # put main here when this is fixed
+  #pull_request:
+  #  paths:
+  #    - "src/**"
+  #    - "tests/**"
+  #    - ".github/**"
+  #  types: [opened, synchronize, reopened]
+
+jobs:
+  run_tests_templates_like:
+    name: "Add new model like template tests"
+    runs-on: ubuntu-22.04
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Install dependencies
+        run: |
+          sudo apt -y update && sudo apt install -y libsndfile1-dev
+
+      - name: Load cached virtual environment
+        uses: actions/cache@v4
+        id: cache
+        with:
+          path: ~/venv/
+          key: v4-tests_model_like-${{ hashFiles('setup.py') }}
+
+      - name: Create virtual environment on cache miss
+        if: steps.cache.outputs.cache-hit != 'true'
+        run: |
+          python -m venv ~/venv && . ~/venv/bin/activate
+          pip install --upgrade pip!=21.3
+          pip install -e .[dev]
+
+      - name: Check transformers location
+        # make `transformers` available as package (required since we use `-e` flag) and check it's indeed from the repo.
+        run: |
+          . ~/venv/bin/activate
+          python setup.py develop
+          transformers_install=$(pip list -e | grep transformers)
+          transformers_install_array=($transformers_install)
+          transformers_loc=${transformers_install_array[-1]}
+          transformers_repo_loc=$(pwd .)
+          if [ "$transformers_loc" != "$transformers_repo_loc" ]; then
+              echo "transformers is from $transformers_loc but it shoud be from $transformers_repo_loc/src."
+              echo "A fix is required. Stop testing."
+              exit 1
+          fi
+
+      - name: Create model files
+        run: |
+          . ~/venv/bin/activate
+          transformers-cli add-new-model-like --config_file tests/fixtures/add_distilbert_like_config.json --path_to_repo .
+          make style
+          make fix-copies
+
+      - name: Run all PyTorch modeling test
+        run: |
+          . ~/venv/bin/activate
+          python -m pytest -n 2 --dist=loadfile -s --make-reports=tests_new_models tests/bert_new/test_modeling_bert_new.py
+
+      - name: Run style changes
+        run: |
+          . ~/venv/bin/activate
+          make style && make quality && make repo-consistency
+
+      - name: Failure short reports
+        if: ${{ always() }}
+        run: cat reports/tests_new_models/failures_short.txt
+
+      - name: Test suite reports artifacts
+        if: ${{ always() }}
+        uses: actions/upload-artifact@v4
+        with:
+          name: run_all_tests_new_models_test_reports
+          path: reports/tests_new_models

.github/workflows/benchmark.yml

@@ -0,0 +1,68 @@
+name: Self-hosted runner (benchmark)
+
+on:
+  push:
+    branches: [main]
+  pull_request:
+    types: [ opened, labeled, reopened, synchronize ]
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.head_ref || github.run_id }}
+  cancel-in-progress: true
+
+env:
+  HF_HOME: /mnt/cache
+
+jobs:
+  benchmark:
+    name: Benchmark
+    runs-on:
+      group: aws-g5-4xlarge-cache
+    if: |
+      (github.event_name == 'pull_request' && contains( github.event.pull_request.labels.*.name, 'run-benchmark') )||
+      (github.event_name == 'push' && github.ref == 'refs/heads/main')
+    container:
+      image: huggingface/transformers-pytorch-gpu
+      options: --gpus all --privileged --ipc host
+    steps:
+      - name: Get repo
+        uses: actions/checkout@v4
+        with:
+          ref: ${{ github.event.pull_request.head.sha || github.sha }}
+
+      - name: Install libpq-dev & psql
+        run: |
+          apt update
+          apt install -y libpq-dev postgresql-client
+
+      - name: Install benchmark script dependencies
+        run: python3 -m pip install -r benchmark/requirements.txt
+
+      - name: Reinstall transformers in edit mode (remove the one installed during docker image build)
+        working-directory: /transformers
+        run: python3 -m pip uninstall -y transformers && python3 -m pip install -e ".[torch]"
+
+      - name: Run database init script
+        run: |
+          psql -f benchmark/init_db.sql
+        env:
+          PGDATABASE: metrics
+          PGHOST: ${{ secrets.TRANSFORMERS_BENCHMARKS_PGHOST }}
+          PGUSER: transformers_benchmarks
+          PGPASSWORD: ${{ secrets.TRANSFORMERS_BENCHMARKS_PGPASSWORD }}
+
+      - name: Run benchmark
+        run: |
+          git config --global --add safe.directory /__w/transformers/transformers
+          if [ "$GITHUB_EVENT_NAME" = "pull_request" ]; then
+            commit_id=$(echo "${{ github.event.pull_request.head.sha }}")
+          elif [ "$GITHUB_EVENT_NAME" = "push" ]; then
+            commit_id=$GITHUB_SHA
+          fi
+          commit_msg=$(git show -s --format=%s | cut -c1-70)
+          python3 benchmark/llama.py "${{ github.head_ref || github.ref_name }}" "$commit_id" "$commit_msg"
+        env:
+          HF_TOKEN: ${{ secrets.HF_HUB_READ_TOKEN }}
+          PGHOST: ${{ secrets.TRANSFORMERS_BENCHMARKS_PGHOST }}
+          PGUSER: transformers_benchmarks
+          PGPASSWORD: ${{ secrets.TRANSFORMERS_BENCHMARKS_PGPASSWORD }}

.github/workflows/build-ci-docker-images.yml

@@ -0,0 +1,77 @@
+name: Build pr ci-docker
+
+on:
+  push:
+    branches:
+      - push-ci-image # for now let's only build on this branch
+  repository_dispatch:
+  workflow_call:
+    inputs:
+      image_postfix:
+        required: true
+        type: string
+  schedule:
+    - cron: "6 0 * * *"
+
+
+concurrency:
+  group: ${{ github.workflow }}
+  cancel-in-progress: true
+
+jobs:
+  build:
+    runs-on: ubuntu-22.04
+
+    if: ${{ contains(github.event.head_commit.message, '[build-ci-image]') || contains(github.event.head_commit.message, '[push-ci-image]') && '!cancelled()' || github.event_name == 'schedule' }}
+
+    strategy:
+      matrix:
+        file: ["quality", "consistency", "custom-tokenizers", "torch-light", "tf-light", "exotic-models", "torch-tf-light", "torch-jax-light", "jax-light", "examples-torch",  "examples-tf"]
+    continue-on-error: true
+
+    steps:
+      -
+        name: Set tag
+        run: |
+              if ${{contains(github.event.head_commit.message, '[build-ci-image]')}}; then
+                  echo "TAG=huggingface/transformers-${{ matrix.file }}:dev" >> "$GITHUB_ENV" 
+                  echo "setting it to DEV!"
+              else
+                  echo "TAG=huggingface/transformers-${{ matrix.file }}" >> "$GITHUB_ENV"
+                  
+              fi
+      -
+        name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+      -
+        name: Check out code
+        uses: actions/checkout@v4
+      -
+        name: Login to DockerHub
+        uses: docker/login-action@v3
+        with:
+          username: ${{ secrets.DOCKERHUB_USERNAME }}
+          password: ${{ secrets.DOCKERHUB_PASSWORD }}
+      -
+        name: Build ${{ matrix.file }}.dockerfile
+        uses: docker/build-push-action@v5
+        with:
+          context: ./docker
+          build-args: |
+            REF=${{ github.sha }}
+          file: "./docker/${{ matrix.file }}.dockerfile"
+          push: ${{ contains(github.event.head_commit.message, 'ci-image]') ||  github.event_name == 'schedule' }}
+          tags: ${{ env.TAG }}
+
+  notify:
+    runs-on: ubuntu-22.04
+    if: ${{ contains(github.event.head_commit.message, '[build-ci-image]') || contains(github.event.head_commit.message, '[push-ci-image]') && '!cancelled()' || github.event_name == 'schedule' }}
+    steps:
+      - name: Post to Slack
+        if: ${{ contains(github.event.head_commit.message, '[push-ci-image]') && github.event_name != 'schedule' }}
+        uses: huggingface/hf-workflows/.github/actions/post-slack@main
+        with:
+          slack_channel: "#transformers-ci-circleci-images"
+          title: 🤗 New docker images for CircleCI are pushed.
+          status: ${{ job.status }}
+          slack_token: ${{ secrets.SLACK_CIFEEDBACK_BOT_TOKEN }}