Merge branch 'main' into feature-disable-no-sync

.circleci/create_circleci_config.py

@@ -57,9 +57,9 @@ class CircleCIJob:
     install_steps: List[str] = None
     marker: Optional[str] = None
     parallelism: Optional[int] = 1
-    pytest_num_workers: int = 8
+    pytest_num_workers: int = 12
     pytest_options: Dict[str, Any] = None
-    resource_class: Optional[str] = "xlarge"
+    resource_class: Optional[str] = "2xlarge"
     tests_to_run: Optional[List[str]] = None
     working_directory: str = "~/transformers"
     # This should be only used for doctest job!
@@ -317,7 +317,7 @@ def job_name(self):
         "pip install -U --upgrade-strategy eager -e git+https://github.com/huggingface/accelerate@main#egg=accelerate",
     ],
     parallelism=1,
-    pytest_num_workers=6,
+    pytest_num_workers=12,
 )
 
 
@@ -353,7 +353,7 @@ def job_name(self):
         "pip install -U --upgrade-strategy eager .[sklearn,torch,testing,sentencepiece,torch-speech,vision,timm,video]",
     ],
     marker="is_pipeline_test",
-    pytest_num_workers=6,
+    pytest_num_workers=12,
 )
 
 

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

@@ -59,7 +59,7 @@ jobs:
 
   latest-torch-deepspeed-docker:
     name: "Latest PyTorch + DeepSpeed"
-    runs-on: ubuntu-22.04
+    runs-on: [intel-cpu, 8-cpu, ci]
     steps:
       - name: Cleanup disk
         run: |
@@ -96,7 +96,7 @@ jobs:
   # Can't build 2 images in a single job `latest-torch-deepspeed-docker` (for `nvcr.io/nvidia`)
   latest-torch-deepspeed-docker-for-push-ci-daily-build:
     name: "Latest PyTorch + DeepSpeed (Push CI - Daily Build)"
-    runs-on: ubuntu-22.04
+    runs-on: [intel-cpu, 8-cpu, ci]
     steps:
       - name: Cleanup disk
         run: |
@@ -138,7 +138,7 @@ jobs:
     name: "Doc builder"
     # Push CI doesn't need this image
     if: inputs.image_postfix != '-push-ci'
-    runs-on: ubuntu-22.04
+    runs-on: [intel-cpu, 8-cpu, ci]
     steps:
       -
         name: Set up Docker Buildx
@@ -164,7 +164,7 @@ jobs:
     name: "Latest PyTorch [dev]"
     # Push CI doesn't need this image
     if: inputs.image_postfix != '-push-ci'
-    runs-on: ubuntu-22.04
+    runs-on: [intel-cpu, 8-cpu, ci]
     steps:
       - name: Cleanup disk
         run: |
@@ -238,7 +238,7 @@ jobs:
     name: "Latest TensorFlow [dev]"
     # Push CI doesn't need this image
     if: inputs.image_postfix != '-push-ci'
-    runs-on: ubuntu-22.04
+    runs-on: [intel-cpu, 8-cpu, ci]
     steps:
       -
         name: Set up Docker Buildx

conftest.py

@@ -30,9 +30,16 @@
     "test_tokenization",
     "test_processor",
     "test_processing",
+    "test_beam_constraints",
+    "test_configuration_utils",
+    "test_data_collator",
+    "test_trainer_callback",
+    "test_trainer_utils",
     "test_feature_extraction",
     "test_image_processing",
     "test_image_processor",
+    "test_image_transforms",
+    "test_optimization",
     "test_retrieval",
     "test_config",
     "test_from_pretrained_no_checkpoint",
@@ -62,6 +69,9 @@
     "test_flax_from_pt_safetensors",
     "ModelTest::test_pipeline_",  # None of the pipeline tests from PipelineTesterMixin (of which XxxModelTest inherits from) are running on device
     "ModelTester::test_pipeline_",
+    "/repo_utils/",
+    "/utils/",
+    "/tools/",
 }
 
 # allow having multiple repository checkouts and not needing to remember to rerun

docker/transformers-all-latest-gpu/Dockerfile

@@ -9,7 +9,7 @@ SHELL ["sh", "-lc"]
 # The following `ARG` are mainly used to specify the versions explicitly & directly in this docker file, and not meant
 # to be used as arguments for docker build (so far).
 
-ARG PYTORCH='2.2.0'
+ARG PYTORCH='2.2.1'
 # (not always a valid torch version)
 ARG INTEL_TORCH_EXT='2.2.0'
 # Example: `cu102`, `cu113`, etc.
@@ -23,21 +23,10 @@ RUN python3 -m pip install --no-cache-dir --upgrade pip
 ARG REF=main
 RUN git clone https://github.com/huggingface/transformers && cd transformers && git checkout $REF
 
-# During switch torch 2.2, we need to move (explicit) torch installation below but keep tf installation here.
-# (otherwise we get `The runner has received a shutdown signal.` whose root cause is unknown but likely disk being full)
-RUN python3 -m pip install --no-cache-dir -U tensorflow==2.13 protobuf==3.20.3 tensorflow_text tensorflow_probability
-
-RUN python3 -m pip install --no-cache-dir -e ./transformers[dev,onnxruntime]
-
-# RUN python3 -m pip uninstall -y torch torchvision torchaudio && python3 -m pip install --no-cache-dir -U torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118
-
-# TODO: Handle these in a python utility script
-RUN [ ${#PYTORCH} -gt 0 -a "$PYTORCH" != "pre" ] && VERSION='torch=='$PYTORCH'.*' ||  VERSION='torch'; echo "export VERSION='$VERSION'" >> ~/.profile
-RUN echo torch=$VERSION
-# `torchvision` and `torchaudio` should be installed along with `torch`, especially for nightly build.
-# Currently, let's just use their latest releases (when `torch` is installed with a release version)
-# TODO: We might need to specify proper versions that work with a specific torch version (especially for past CI).
-RUN [ "$PYTORCH" != "pre" ] && python3 -m pip install --no-cache-dir -U $VERSION torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/$CUDA || python3 -m pip install --no-cache-dir -U --pre torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/nightly/$CUDA
+# 1. Put several commands in a single `RUN` to avoid image/layer exporting issue. Could be revised in the future.
+# 2. Regarding `torch` part, We might need to specify proper versions for `torchvision` and `torchaudio`.
+#    Currently, let's not bother to specify their versions explicitly (so installed with their latest release versions).
+RUN python3 -m pip install --no-cache-dir -U tensorflow==2.13 protobuf==3.20.3 tensorflow_text tensorflow_probability && python3 -m pip install --no-cache-dir -e ./transformers[dev,onnxruntime] && [ ${#PYTORCH} -gt 0 -a "$PYTORCH" != "pre" ] && VERSION='torch=='$PYTORCH'.*' ||  VERSION='torch'; echo "export VERSION='$VERSION'" >> ~/.profile && echo torch=$VERSION && [ "$PYTORCH" != "pre" ] && python3 -m pip install --no-cache-dir -U $VERSION torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/$CUDA || python3 -m pip install --no-cache-dir -U --pre torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/nightly/$CUDA
 
 RUN python3 -m pip uninstall -y flax jax
 

docker/transformers-quantization-latest-gpu/Dockerfile

@@ -45,6 +45,9 @@ RUN python3 -m pip install --no-cache-dir aqlm[gpu]==1.0.2
 # Add autoawq for quantization testing
 RUN python3 -m pip install --no-cache-dir https://github.com/casper-hansen/AutoAWQ/releases/download/v0.2.0/autoawq-0.2.0+cu118-cp38-cp38-linux_x86_64.whl
 
+# Add quanto for quantization testing
+RUN python3 -m pip install --no-cache-dir quanto
+
 # When installing in editable mode, `transformers` is not recognized as a package.
 # this line must be added in order for python to be aware of transformers.
 RUN cd transformers && python3 setup.py develop

docs/source/en/_toctree.yml

@@ -134,7 +134,7 @@
   - local: custom_tools
     title: Custom Tools and Prompts
   - local: troubleshooting
-    title: Troubleshoot  
+    title: Troubleshoot
   - local: hf_quantizer
     title: Contribute new quantization method
   title: Developer guides
@@ -310,6 +310,8 @@
         title: CodeGen
       - local: model_doc/code_llama
         title: CodeLlama
+      - local: model_doc/cohere
+        title: Cohere
       - local: model_doc/convbert
         title: ConvBERT
       - local: model_doc/cpm
@@ -593,6 +595,8 @@
         title: SegFormer
       - local: model_doc/seggpt
         title: SegGpt
+      - local: model_doc/superpoint
+        title: SuperPoint
       - local: model_doc/swiftformer
         title: SwiftFormer
       - local: model_doc/swin
@@ -640,6 +644,8 @@
         title: MMS
       - local: model_doc/musicgen
         title: MusicGen
+      - local: model_doc/musicgen_melody
+        title: MusicGen Melody
       - local: model_doc/pop2piano
         title: Pop2Piano
       - local: model_doc/seamless_m4t
@@ -689,7 +695,7 @@
         title: VideoMAE
       - local: model_doc/vivit
         title: ViViT
-      title: Video models 
+      title: Video models
     - isExpanded: false
       sections:
       - local: model_doc/align
@@ -742,6 +748,8 @@
         title: LiLT
       - local: model_doc/llava
         title: Llava
+      - local: model_doc/llava_next
+        title: LLaVA-NeXT
       - local: model_doc/lxmert
         title: LXMERT
       - local: model_doc/matcha

docs/source/en/custom_tools.md

@@ -51,10 +51,10 @@ seemingly giving the agent some kind of memory.
 Let's take a closer look at how the prompt is structured to understand how it can be best customized.
 The prompt is structured broadly into four parts.
 
-- 1. Introduction: how the agent should behave, explanation of the concept of tools.
-- 2. Description of all the tools. This is defined by a `<<all_tools>>` token that is dynamically replaced at runtime with the tools defined/chosen by the user.
-- 3. A set of examples of tasks and their solution
-- 4. Current example, and request for solution.
+1. Introduction: how the agent should behave, explanation of the concept of tools.
+2. Description of all the tools. This is defined by a `<<all_tools>>` token that is dynamically replaced at runtime with the tools defined/chosen by the user.
+3. A set of examples of tasks and their solution
+4. Current example, and request for solution.
 
 To better understand each part, let's look at a shortened version of how the `run` prompt can look like:
 
@@ -301,7 +301,7 @@ and description and try refining your task request with it.
 ### Customizing the tool descriptions
 
 As we've seen before the agent has access to each of the tools' names and descriptions. The base tools 
-should have very precise names and descriptions, however, you might find that it could help to change the 
+should have very precise names and descriptions, however, you might find that it could help to change 
 the description or name of a tool for your specific use case. This might become especially important 
 when you've added multiple tools that are very similar or if you want to use your agent only for a certain 
 domain, *e.g.* image generation and transformations.

docs/source/en/index.md

@@ -95,6 +95,7 @@ Flax), PyTorch, and/or TensorFlow.
 |                          [CLVP](model_doc/clvp)                          |       ✅        |         ❌         |      ❌      |
 |                       [CodeGen](model_doc/codegen)                       |       ✅        |         ❌         |      ❌      |
 |                    [CodeLlama](model_doc/code_llama)                     |       ✅        |         ❌         |      ✅      |
+|                        [Cohere](model_doc/cohere)                        |       ✅        |         ❌         |      ❌      |
 |              [Conditional DETR](model_doc/conditional_detr)              |       ✅        |         ❌         |      ❌      |
 |                      [ConvBERT](model_doc/convbert)                      |       ✅        |         ✅         |      ❌      |
 |                      [ConvNeXT](model_doc/convnext)                      |       ✅        |         ✅         |      ❌      |
@@ -173,6 +174,7 @@ Flax), PyTorch, and/or TensorFlow.
 |                         [LLaMA](model_doc/llama)                         |       ✅        |         ❌         |      ✅      |
 |                        [Llama2](model_doc/llama2)                        |       ✅        |         ❌         |      ✅      |
 |                         [LLaVa](model_doc/llava)                         |       ✅        |         ❌         |      ❌      |
+|                    [LLaVA-NeXT](model_doc/llava_next)                    |       ✅        |         ❌         |      ❌      |
 |                    [Longformer](model_doc/longformer)                    |       ✅        |         ✅         |      ❌      |
 |                        [LongT5](model_doc/longt5)                        |       ✅        |         ❌         |      ✅      |
 |                          [LUKE](model_doc/luke)                          |       ✅        |         ❌         |      ❌      |
@@ -206,6 +208,7 @@ Flax), PyTorch, and/or TensorFlow.
 |                           [MRA](model_doc/mra)                           |       ✅        |         ❌         |      ❌      |
 |                           [MT5](model_doc/mt5)                           |       ✅        |         ✅         |      ✅      |
 |                      [MusicGen](model_doc/musicgen)                      |       ✅        |         ❌         |      ❌      |
+|               [MusicGen Melody](model_doc/musicgen_melody)               |       ✅        |         ❌         |      ❌      |
 |                           [MVP](model_doc/mvp)                           |       ✅        |         ❌         |      ❌      |
 |                           [NAT](model_doc/nat)                           |       ✅        |         ❌         |      ❌      |
 |                         [Nezha](model_doc/nezha)                         |       ✅        |         ❌         |      ❌      |
@@ -264,6 +267,7 @@ Flax), PyTorch, and/or TensorFlow.
 |                   [SqueezeBERT](model_doc/squeezebert)                   |       ✅        |         ❌         |      ❌      |
 |                      [StableLm](model_doc/stablelm)                      |       ✅        |         ❌         |      ❌      |
 |                    [Starcoder2](model_doc/starcoder2)                    |       ✅        |         ❌         |      ❌      |
+|                    [SuperPoint](model_doc/superpoint)                    |       ✅        |         ❌         |      ❌      |
 |                   [SwiftFormer](model_doc/swiftformer)                   |       ✅        |         ❌         |      ❌      |
 |                    [Swin Transformer](model_doc/swin)                    |       ✅        |         ✅         |      ❌      |
 |                 [Swin Transformer V2](model_doc/swinv2)                  |       ✅        |         ❌         |      ❌      |

docs/source/en/main_classes/quantization.md

@@ -26,6 +26,10 @@ Learn how to quantize models in the [Quantization](../quantization) guide.
 
 </Tip>
 
+## QuantoConfig
+
+[[autodoc]] QuantoConfig
+
 ## AqlmConfig
 
 [[autodoc]] AqlmConfig

docs/source/en/model_doc/auto.md

@@ -250,6 +250,10 @@ The following auto classes are available for the following computer vision tasks
 
 [[autodoc]] AutoModelForVideoClassification
 
+### AutoModelForKeypointDetection
+
+[[autodoc]] AutoModelForKeypointDetection
+
 ### AutoModelForMaskedImageModeling
 
 [[autodoc]] AutoModelForMaskedImageModeling