Merge branch 'main' into plr_fixes

Author: cyyever

.circleci/create_circleci_config.py

@@ -109,7 +109,7 @@ def __post_init__(self):
                 self.docker_image[0]["image"] = f"{self.docker_image[0]['image']}:dev"
             print(f"Using {self.docker_image} docker image")
         if self.install_steps is None:
-            self.install_steps = ["uv venv && uv pip install ."]
+            self.install_steps = ["uv pip install ."]
         if self.pytest_options is None:
             self.pytest_options = {}
         if isinstance(self.tests_to_run, str):
@@ -213,7 +213,7 @@ def job_name(self):
     docker_image=[{"image": "huggingface/transformers-torch-light"}],
     # networkx==3.3 (after #36957) cause some issues
     # TODO: remove this once it works directly
-    install_steps=["uv venv && uv pip install ."],
+    install_steps=["uv pip install ."],
     marker="generate",
     parallelism=6,
 )
@@ -250,7 +250,7 @@ def job_name(self):
     additional_env={"OMP_NUM_THREADS": 8},
     docker_image=[{"image":"huggingface/transformers-examples-torch"}],
     # TODO @ArthurZucker remove this once docker is easier to build
-    install_steps=["uv venv && uv pip install . && uv pip install -r examples/pytorch/_tests_requirements.txt"],
+    install_steps=["uv pip install . && uv pip install -r examples/pytorch/_tests_requirements.txt"],
     pytest_num_workers=4,
 )
 
@@ -259,7 +259,7 @@ def job_name(self):
     additional_env={"HUGGINGFACE_CO_STAGING": True},
     docker_image=[{"image":"huggingface/transformers-torch-light"}],
     install_steps=[
-        'uv venv && uv pip install .',
+        'uv pip install .',
         'git config --global user.email "[email protected]"',
         'git config --global user.name "ci"',
     ],
@@ -273,7 +273,6 @@ def job_name(self):
     "onnx",
     docker_image=[{"image":"huggingface/transformers-torch-tf-light"}],
     install_steps=[
-        "uv venv",
         "uv pip install .[testing,sentencepiece,onnxruntime,vision,rjieba]",
     ],
     pytest_options={"k onnx": None},
@@ -303,7 +302,7 @@ def job_name(self):
     docker_image=[{"image": "huggingface/transformers-torch-light"}],
     # networkx==3.3 (after #36957) cause some issues
     # TODO: remove this once it works directly
-    install_steps=["uv venv && uv pip install .[serving]"],
+    install_steps=["uv pip install .[serving]"],
     marker="not generate",
     parallelism=6,
 )
@@ -321,7 +320,7 @@ def job_name(self):
     additional_env={"TRANSFORMERS_VERBOSITY": "error", "DATASETS_VERBOSITY": "error", "SKIP_CUDA_DOCTEST": "1"},
     install_steps=[
         # Add an empty file to keep the test step running correctly even no file is selected to be tested.
-        "uv venv && pip install .",
+        "uv pip install .",
         "touch dummy.py",
         command,
         "cat pr_documentation_tests_temp.txt",

.github/workflows/self-scheduled-amd-mi355-caller.yml

@@ -0,0 +1,63 @@
+name: Self-hosted runner scale set (AMD mi355 scheduled CI caller)
+
+# Note: For every job in this workflow, the name of the runner scale set is finalized in the runner yaml i.e. huggingface/hf-workflows/.github/workflows/transformers_amd_ci_scheduled_arc_scale_set.yaml
+# For example, 1gpu : amd-mi355-ci-1gpu
+#              2gpu : amd-mi355-ci-2gpu
+
+on:
+  workflow_run:
+    workflows: ["Self-hosted runner (AMD scheduled CI caller)"]
+    branches: ["main"]
+    types: [completed]
+  push:
+    branches:
+      - run_amd_scheduled_ci_caller*
+
+jobs:
+  model-ci:
+    name: Model CI
+    uses: huggingface/hf-workflows/.github/workflows/transformers_amd_ci_scheduled_arc_scale_set.yaml@main
+    with:
+      job: run_models_gpu
+      slack_report_channel: "#amd-hf-ci"
+      runner_scale_set: amd-mi355-ci
+      docker: huggingface/transformers-pytorch-amd-gpu
+      ci_event: Scheduled CI (AMD) - mi355
+      report_repo_id: optimum-amd/transformers_daily_ci
+    secrets: inherit
+
+  torch-pipeline:
+    name: Torch pipeline CI
+    uses: huggingface/hf-workflows/.github/workflows/transformers_amd_ci_scheduled_arc_scale_set.yaml@main
+    with:
+      job: run_pipelines_torch_gpu
+      slack_report_channel: "#amd-hf-ci"
+      runner_scale_set: amd-mi355-ci
+      docker: huggingface/transformers-pytorch-amd-gpu
+      ci_event: Scheduled CI (AMD) - mi355
+      report_repo_id: optimum-amd/transformers_daily_ci
+    secrets: inherit
+
+  example-ci:
+    name: Example CI
+    uses: huggingface/hf-workflows/.github/workflows/transformers_amd_ci_scheduled_arc_scale_set.yaml@main
+    with:
+      job: run_examples_gpu
+      slack_report_channel: "#amd-hf-ci"
+      runner_scale_set: amd-mi355-ci
+      docker: huggingface/transformers-pytorch-amd-gpu
+      ci_event: Scheduled CI (AMD) - mi355
+      report_repo_id: optimum-amd/transformers_daily_ci
+    secrets: inherit
+
+  deepspeed-ci:
+    name: DeepSpeed CI
+    uses: huggingface/hf-workflows/.github/workflows/transformers_amd_ci_scheduled_arc_scale_set.yaml@main
+    with:
+      job: run_torch_cuda_extensions_gpu
+      slack_report_channel: "#amd-hf-ci"
+      runner_scale_set: amd-mi355-ci
+      docker: huggingface/transformers-pytorch-deepspeed-amd-gpu
+      ci_event: Scheduled CI (AMD) - mi355
+      report_repo_id: optimum-amd/transformers_daily_ci
+    secrets: inherit

conftest.py

@@ -83,6 +83,8 @@ def pytest_configure(config):
     config.addinivalue_line("markers", "is_staging_test: mark test to run only in the staging environment")
     config.addinivalue_line("markers", "accelerate_tests: mark test that require accelerate")
     config.addinivalue_line("markers", "not_device_test: mark the tests always running on cpu")
+    config.addinivalue_line("markers", "torch_compile_test: mark test which tests torch compile functionality")
+    config.addinivalue_line("markers", "torch_export_test: mark test which tests torch export functionality")
 
 
 def pytest_collection_modifyitems(items):

docker/consistency.dockerfile

@@ -4,7 +4,7 @@ USER root
 ARG REF=main
 RUN apt-get update && apt-get install -y time git g++ pkg-config make git-lfs
 ENV UV_PYTHON=/usr/local/bin/python
-RUN pip install uv && uv venv && uv pip install --no-cache-dir -U pip setuptools GitPython
+RUN pip install uv && uv pip install --no-cache-dir -U pip setuptools GitPython
 RUN uv pip install --no-cache-dir --upgrade 'torch' 'torchaudio' 'torchvision' --index-url https://download.pytorch.org/whl/cpu
 # tensorflow pin matching setup.py
 RUN uv pip install --no-cache-dir pypi-kenlm

docker/custom-tokenizers.dockerfile

@@ -4,7 +4,7 @@ ARG REF=main
 USER root
 RUN apt-get update && apt-get install -y libsndfile1-dev espeak-ng time git cmake wget xz-utils build-essential g++5 libprotobuf-dev protobuf-compiler
 ENV UV_PYTHON=/usr/local/bin/python
-RUN pip --no-cache-dir install uv && uv venv && uv pip install --no-cache-dir -U pip setuptools
+RUN pip --no-cache-dir install uv && uv pip install --no-cache-dir -U pip setuptools
 
 RUN wget https://github.com/ku-nlp/jumanpp/releases/download/v2.0.0-rc3/jumanpp-2.0.0-rc3.tar.xz
 RUN tar xvf jumanpp-2.0.0-rc3.tar.xz
@@ -20,7 +20,7 @@ RUN uv pip install --no-cache --upgrade 'torch' --index-url https://download.pyt
 RUN uv pip install --no-cache-dir  --no-deps accelerate --extra-index-url https://download.pytorch.org/whl/cpu
 RUN uv pip install  --no-cache-dir "git+https://github.com/huggingface/transformers.git@${REF}#egg=transformers[ja,testing,sentencepiece,jieba,spacy,ftfy,rjieba]" unidic unidic-lite
 # spacy is not used so not tested. Causes to failures. TODO fix later
-RUN python3 -m unidic download
+RUN uv run python -m unidic download
 RUN uv pip uninstall transformers
 
 RUN apt-get clean && rm -rf /var/lib/apt/lists/*

docker/examples-tf.dockerfile

@@ -5,7 +5,7 @@ USER root
 RUN apt-get update && apt-get install -y libsndfile1-dev espeak-ng time git
 RUN apt-get install -y g++ cmake
 ENV UV_PYTHON=/usr/local/bin/python
-RUN pip --no-cache-dir install uv && uv venv
+RUN pip --no-cache-dir install uv
 RUN uv pip install --no-cache-dir -U pip setuptools albumentations seqeval
 RUN uv pip install  --upgrade --no-cache-dir "git+https://github.com/huggingface/transformers.git@${REF}#egg=transformers[tf-cpu,sklearn,testing,sentencepiece,tf-speech,vision]"
 RUN uv pip install --no-cache-dir  "protobuf==3.20.3"

docker/examples-torch.dockerfile

@@ -4,7 +4,7 @@ ARG REF=main
 USER root
 RUN apt-get update &&  apt-get install -y --no-install-recommends libsndfile1-dev espeak-ng time git g++ cmake pkg-config openssh-client git ffmpeg
 ENV UV_PYTHON=/usr/local/bin/python
-RUN pip --no-cache-dir install uv && uv venv && uv pip install --no-cache-dir -U pip setuptools
+RUN pip --no-cache-dir install uv && uv pip install --no-cache-dir -U pip setuptools
 RUN uv pip install --no-cache-dir 'torch' 'torchaudio' 'torchvision' 'torchcodec' --index-url https://download.pytorch.org/whl/cpu
 RUN uv pip install --no-deps timm accelerate --extra-index-url https://download.pytorch.org/whl/cpu
 RUN uv pip install --no-cache-dir librosa "git+https://github.com/huggingface/transformers.git@${REF}#egg=transformers[sklearn,sentencepiece,vision,testing]" seqeval albumentations jiwer

docker/exotic-models.dockerfile

@@ -4,10 +4,10 @@ ARG REF=main
 USER root
 RUN apt-get update && apt-get install -y libsndfile1-dev espeak-ng time git libgl1-mesa-glx libgl1 g++ tesseract-ocr
 ENV UV_PYTHON=/usr/local/bin/python
-RUN pip --no-cache-dir install uv &&  uv venv && uv pip install --no-cache-dir -U pip setuptools
+RUN pip --no-cache-dir install uv && uv pip install --no-cache-dir -U pip setuptools
 RUN uv pip install --no-cache-dir 'torch' 'torchaudio' 'torchvision' --index-url https://download.pytorch.org/whl/cpu
 RUN uv pip install --no-cache-dir  --no-deps timm accelerate
-RUN pip install -U --upgrade-strategy eager --no-cache-dir pytesseract python-Levenshtein opencv-python nltk
+RUN uv pip install -U --upgrade-strategy eager --no-cache-dir pytesseract python-Levenshtein opencv-python nltk
 # RUN uv pip install --no-cache-dir natten==0.15.1+torch210cpu -f https://shi-labs.com/natten/wheels
 RUN uv pip install  --no-cache-dir "git+https://github.com/huggingface/transformers.git@${REF}#egg=transformers[testing, vision]" 'scikit-learn' 'torch-stft' 'nose'  'dataset'
 # RUN git clone https://github.com/facebookresearch/detectron2.git

docker/jax-light.dockerfile

@@ -4,7 +4,7 @@ ARG REF=main
 USER root
 RUN apt-get update && apt-get install -y libsndfile1-dev espeak-ng time git g++ cmake
 ENV UV_PYTHON=/usr/local/bin/python
-RUN pip --no-cache-dir install uv &&  uv venv && uv pip install --no-cache-dir -U pip setuptools
+RUN pip --no-cache-dir install uv && uv pip install --no-cache-dir -U pip setuptools
 RUN uv pip install --no-cache-dir "scipy<1.13" "git+https://github.com/huggingface/transformers.git@${REF}#egg=transformers[flax,testing,sentencepiece,flax-speech,vision]"
 RUN uv pip uninstall transformers
 RUN apt-get clean && rm -rf /var/lib/apt/lists/* && apt-get autoremove && apt-get autoclean

docker/pipeline-tf.dockerfile

@@ -4,7 +4,7 @@ ARG REF=main
 USER root
 RUN apt-get update && apt-get install -y libsndfile1-dev espeak-ng time git cmake g++
 ENV UV_PYTHON=/usr/local/bin/python
-RUN pip --no-cache-dir install uv && uv venv && uv pip install --no-cache-dir -U pip setuptools
+RUN pip --no-cache-dir install uv && uv pip install --no-cache-dir -U pip setuptools
 RUN uv pip install --no-cache-dir "git+https://github.com/huggingface/transformers.git@${REF}#egg=transformers[sklearn,tf-cpu,testing,sentencepiece,tf-speech,vision]"
 RUN uv pip install --no-cache-dir  "protobuf==3.20.3" tensorflow_probability
 RUN apt-get clean && rm -rf /var/lib/apt/lists/*