update

.github/workflows/pr_tests.yml

@@ -92,9 +92,8 @@ jobs:
             runner: aws-general-8-plus
             image: diffusers/diffusers-pytorch-cpu
             report: torch_example_cpu
-        transformers_version: ["main"]
 
-    name: ${{ matrix.config.name }} (transformers ${{ matrix.transformers_version }})
+    name: ${{ matrix.config.name }}
 
     runs-on:
       group: ${{ matrix.config.runner }}
@@ -156,7 +155,7 @@ jobs:
       if: ${{ always() }}
       uses: actions/upload-artifact@v6
       with:
-        name: pr_${{ matrix.config.framework }}_${{ matrix.config.report }}_transformers_${{ matrix.transformers_version }}_test_reports
+        name: pr_${{ matrix.config.framework }}_${{ matrix.config.report }}_test_reports
         path: reports
 
   run_staging_tests:
@@ -221,10 +220,8 @@ jobs:
     needs: [check_code_quality, check_repository_consistency]
     strategy:
       fail-fast: false
-      matrix:
-        transformers_version: ["main"]
 
-    name: LoRA tests with PEFT main (transformers ${{ matrix.transformers_version }})
+    name: LoRA tests with PEFT main
 
     runs-on:
       group: aws-general-8-plus
@@ -250,12 +247,9 @@ jobs:
         uv pip install -U peft@git+https://github.com/huggingface/peft.git --no-deps
         uv pip install -U tokenizers
         uv pip uninstall accelerate && uv pip install -U accelerate@git+https://github.com/huggingface/accelerate.git --no-deps
-        if [ "${{ matrix.transformers_version }}" = "main" ]; then
-          uv pip uninstall transformers huggingface_hub && uv pip install --prerelease allow -U transformers@git+https://github.com/huggingface/transformers.git
-        else
-          uv pip uninstall transformers huggingface_hub && uv pip install transformers==${{ matrix.transformers_version }}
-        fi
-        
+        #uv pip uninstall transformers huggingface_hub && uv pip install --prerelease allow -U transformers@git+https://github.com/huggingface/transformers.git
+        uv pip uninstall transformers huggingface_hub && uv pip install transformers==4.57.1
+
     - name: Environment
       run: |
         python utils/print_env.py
@@ -281,6 +275,6 @@ jobs:
       if: ${{ always() }}
       uses: actions/upload-artifact@v6
       with:
-        name: pr_lora_transformers_${{ matrix.transformers_version }}_test_reports
+        name: pr_main_test_reports
         path: reports
 

.github/workflows/pr_tests_gpu.yml

@@ -14,7 +14,6 @@ on:
       - "tests/pipelines/test_pipelines_common.py"
       - "tests/models/test_modeling_common.py"
       - "examples/**/*.py"
-      - ".github/**.yml"
   workflow_dispatch:
 
 concurrency:
@@ -107,14 +106,13 @@ jobs:
           path: reports
 
   torch_pipelines_cuda_tests:
-    name: Torch Pipelines CUDA Tests (transformers ${{ matrix.transformers_version }})
+    name: Torch Pipelines CUDA Tests
     needs: setup_torch_cuda_pipeline_matrix
     strategy:
       fail-fast: false
       max-parallel: 8
       matrix:
         module: ${{ fromJson(needs.setup_torch_cuda_pipeline_matrix.outputs.pipeline_test_matrix) }}
-        transformers_version: ["main"]
     runs-on:
       group: aws-g4dn-2xlarge
     container:
@@ -133,12 +131,8 @@ jobs:
         run: |
           uv pip install -e ".[quality]"
           uv pip uninstall accelerate && uv pip install -U accelerate@git+https://github.com/huggingface/accelerate.git
-          if [ "${{ matrix.transformers_version }}" = "main" ]; then
-            uv pip uninstall transformers huggingface_hub && uv pip install --prerelease allow -U transformers@git+https://github.com/huggingface/transformers.git
-          else
-            uv pip uninstall transformers huggingface_hub && uv pip install transformers==${{ matrix.transformers_version }}
-          fi
-          
+          #uv pip uninstall transformers huggingface_hub && uv pip install --prerelease allow -U transformers@git+https://github.com/huggingface/transformers.git
+          uv pip uninstall transformers huggingface_hub && uv pip install transformers==4.57.1
 
       - name: Environment
         run: |
@@ -178,11 +172,11 @@ jobs:
         if: ${{ always() }}
         uses: actions/upload-artifact@v6
         with:
-          name: pipeline_${{ matrix.module }}_transformers_${{ matrix.transformers_version }}_test_reports
+          name: pipeline_${{ matrix.module }}_test_reports
           path: reports
 
   torch_cuda_tests:
-    name: Torch CUDA Tests (transformers ${{ matrix.transformers_version }})
+    name: Torch CUDA Tests
     needs: [check_code_quality, check_repository_consistency]
     runs-on:
       group: aws-g4dn-2xlarge
@@ -197,7 +191,6 @@ jobs:
       max-parallel: 4
       matrix:
         module: [models, schedulers, lora, others]
-        transformers_version: ["main"]
     steps:
     - name: Checkout diffusers
       uses: actions/checkout@v6
@@ -209,12 +202,8 @@ jobs:
         uv pip install -e ".[quality]"
         uv pip install peft@git+https://github.com/huggingface/peft.git
         uv pip uninstall accelerate && uv pip install -U accelerate@git+https://github.com/huggingface/accelerate.git
-        if [ "${{ matrix.transformers_version }}" = "main" ]; then
-          uv pip uninstall transformers huggingface_hub && uv pip install --prerelease allow -U transformers@git+https://github.com/huggingface/transformers.git
-        else
-          uv pip uninstall transformers huggingface_hub && uv pip install transformers==${{ matrix.transformers_version }}
-        fi
-        
+        #uv pip uninstall transformers huggingface_hub && uv pip install --prerelease allow -U transformers@git+https://github.com/huggingface/transformers.git
+        uv pip uninstall transformers huggingface_hub && uv pip install transformers==4.57.1
 
     - name: Environment
       run: |
@@ -252,16 +241,12 @@ jobs:
       if: ${{ always() }}
       uses: actions/upload-artifact@v6
       with:
-        name: torch_cuda_test_reports_${{ matrix.module }}_transformers_${{ matrix.transformers_version }}
+        name: torch_cuda_test_reports_${{ matrix.module }}
         path: reports
 
   run_examples_tests:
-    name: Examples PyTorch CUDA tests on Ubuntu (transformers ${{ matrix.transformers_version }})
+    name: Examples PyTorch CUDA tests on Ubuntu
     needs: [check_code_quality, check_repository_consistency]
-    strategy:
-      fail-fast: false
-      matrix:
-        transformers_version: ["main"]
     runs-on:
       group: aws-g4dn-2xlarge
 
@@ -279,11 +264,8 @@ jobs:
         nvidia-smi
     - name: Install dependencies
       run: |
-        if [ "${{ matrix.transformers_version }}" = "main" ]; then
-          uv pip uninstall transformers huggingface_hub && uv pip install --prerelease allow -U transformers@git+https://github.com/huggingface/transformers.git
-        else
-          uv pip uninstall transformers huggingface_hub && uv pip install transformers==${{ matrix.transformers_version }}
-        fi
+        #uv pip uninstall transformers huggingface_hub && uv pip install --prerelease allow -U transformers@git+https://github.com/huggingface/transformers.git
+        uv pip uninstall transformers huggingface_hub && uv pip install transformers==4.57.1
         uv pip install -e ".[quality,training]"
 
     - name: Environment
@@ -307,6 +289,6 @@ jobs:
       if: ${{ always() }}
       uses: actions/upload-artifact@v6
       with:
-        name: examples_transformers_${{ matrix.transformers_version }}_test_reports
+        name: examples_test_reports
         path: reports
 

examples/custom_diffusion/test_custom_diffusion.py

@@ -17,9 +17,6 @@ import logging
 import os
 import sys
 import tempfile
-import unittest
-
-from diffusers.utils import is_transformers_version
 
 
 sys.path.append("..")
@@ -33,7 +30,6 @@ stream_handler = logging.StreamHandler(sys.stdout)
 logger.addHandler(stream_handler)
 
 
-@unittest.skipIf(is_transformers_version(">=", "4.57.5"), "Size mismatch")
 class CustomDiffusion(ExamplesTestsAccelerate):
     def test_custom_diffusion(self):
         with tempfile.TemporaryDirectory() as tmpdir:

src/diffusers/hooks/_common.py

@@ -48,7 +48,6 @@ _GO_LC_SUPPORTED_PYTORCH_LAYERS = (
     torch.nn.ConvTranspose2d,
     torch.nn.ConvTranspose3d,
     torch.nn.Linear,
-    torch.nn.Embedding,
     # TODO(aryan): look into torch.nn.LayerNorm, torch.nn.GroupNorm later, seems to be causing some issues with CogVideoX
     # because of double invocation of the same norm layer in CogVideoXLayerNorm
 )

src/diffusers/loaders/textual_inversion.py

@@ -22,12 +22,7 @@ from tokenizers import Tokenizer as TokenizerFast
 from torch import nn
 
 from ..models.modeling_utils import load_state_dict
-from ..utils import (
-    _get_model_file,
-    is_accelerate_available,
-    is_transformers_available,
-    logging,
-)
+from ..utils import _get_model_file, is_accelerate_available, is_transformers_available, logging
 
 
 if is_transformers_available():

src/diffusers/modular_pipelines/modular_pipeline.py

@@ -14,6 +14,7 @@
 import importlib
 import inspect
 import os
+import sys
 import traceback
 import warnings
 from collections import OrderedDict
@@ -28,10 +29,16 @@ from tqdm.auto import tqdm
 from typing_extensions import Self
 
 from ..configuration_utils import ConfigMixin, FrozenDict
-from ..pipelines.pipeline_loading_utils import _fetch_class_library_tuple, simple_get_class_obj
+from ..pipelines.pipeline_loading_utils import (
+    LOADABLE_CLASSES,
+    _fetch_class_library_tuple,
+    _unwrap_model,
+    simple_get_class_obj,
+)
 from ..utils import PushToHubMixin, is_accelerate_available, logging
 from ..utils.dynamic_modules_utils import get_class_from_dynamic_module, resolve_trust_remote_code
 from ..utils.hub_utils import load_or_create_model_card, populate_model_card
+from ..utils.torch_utils import is_compiled_module
 from .components_manager import ComponentsManager
 from .modular_pipeline_utils import (
     MODULAR_MODEL_CARD_TEMPLATE,
@@ -1819,29 +1826,111 @@ class ModularPipeline(ConfigMixin, PushToHubMixin):
         )
         return pipeline
 
-    def save_pretrained(self, save_directory: str | os.PathLike, push_to_hub: bool = False, **kwargs):
+    def save_pretrained(
+        self,
+        save_directory: str | os.PathLike,
+        safe_serialization: bool = True,
+        variant: str | None = None,
+        max_shard_size: int | str | None = None,
+        push_to_hub: bool = False,
+        **kwargs,
+    ):
         """
-        Save the pipeline to a directory. It does not save components, you need to save them separately.
+        Save the pipeline and all its components to a directory, so that it can be re-loaded using the
+        [`~ModularPipeline.from_pretrained`] class method.
 
         Args:
             save_directory (`str` or `os.PathLike`):
-                Path to the directory where the pipeline will be saved.
-            push_to_hub (`bool`, optional):
-                Whether to push the pipeline to the huggingface hub.
-            **kwargs: Additional arguments passed to `save_config()` method
+                Directory to save the pipeline to. Will be created if it doesn't exist.
+            safe_serialization (`bool`, *optional*, defaults to `True`):
+                Whether to save the model using `safetensors` or the traditional PyTorch way with `pickle`.
+            variant (`str`, *optional*):
+                If specified, weights are saved in the format `pytorch_model.<variant>.bin`.
+            max_shard_size (`int` or `str`, defaults to `None`):
+                The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size
+                lower than this size. If expressed as a string, needs to be digits followed by a unit (like `"5GB"`).
+                If expressed as an integer, the unit is bytes.
+            push_to_hub (`bool`, *optional*, defaults to `False`):
+                Whether to push the pipeline to the Hugging Face model hub after saving it.
+            **kwargs: Additional keyword arguments passed along to the push to hub method.
         """
+        overwrite_modular_index = kwargs.pop("overwrite_modular_index", False)
+        repo_id = kwargs.pop("repo_id", save_directory.split(os.path.sep)[-1])
+
+        for component_name, component_spec in self._component_specs.items():
+            sub_model = getattr(self, component_name, None)
+            if sub_model is None:
+                continue
+
+            model_cls = sub_model.__class__
+            if is_compiled_module(sub_model):
+                sub_model = _unwrap_model(sub_model)
+                model_cls = sub_model.__class__
+
+            save_method_name = None
+            for library_name, library_classes in LOADABLE_CLASSES.items():
+                if library_name in sys.modules:
+                    library = importlib.import_module(library_name)
+                else:
+                    logger.info(
+                        f"{library_name} is not installed. Cannot save {component_name} as {library_classes} from {library_name}"
+                    )
+                    continue
+
+                for base_class, save_load_methods in library_classes.items():
+                    class_candidate = getattr(library, base_class, None)
+                    if class_candidate is not None and issubclass(model_cls, class_candidate):
+                        save_method_name = save_load_methods[0]
+                        break
+                if save_method_name is not None:
+                    break
+
+            if save_method_name is None:
+                logger.warning(f"self.{component_name}={sub_model} of type {type(sub_model)} cannot be saved.")
+                continue
+
+            save_method = getattr(sub_model, save_method_name)
+            save_method_signature = inspect.signature(save_method)
+            save_method_accept_safe = "safe_serialization" in save_method_signature.parameters
+            save_method_accept_variant = "variant" in save_method_signature.parameters
+            save_method_accept_max_shard_size = "max_shard_size" in save_method_signature.parameters
+
+            save_kwargs = {}
+            if save_method_accept_safe:
+                save_kwargs["safe_serialization"] = safe_serialization
+            if save_method_accept_variant:
+                save_kwargs["variant"] = variant
+            if save_method_accept_max_shard_size and max_shard_size is not None:
+                save_kwargs["max_shard_size"] = max_shard_size
+
+            save_method(os.path.join(save_directory, component_name), **save_kwargs)
+
         if push_to_hub:
             commit_message = kwargs.pop("commit_message", None)
             private = kwargs.pop("private", None)
             create_pr = kwargs.pop("create_pr", False)
             token = kwargs.pop("token", None)
-            repo_id = kwargs.pop("repo_id", save_directory.split(os.path.sep)[-1])
             repo_id = create_repo(repo_id, exist_ok=True, private=private, token=token).repo_id
 
-            # Generate modular pipeline card content
-            card_content = generate_modular_model_card_content(self.blocks)
+        if overwrite_modular_index:
+            for component_name, component_spec in self._component_specs.items():
+                if component_spec.default_creation_method != "from_pretrained":
+                    continue
+                if component_name not in self.config:
+                    continue
 
-            # Create a new empty model card and eventually tag it
+                library, class_name, component_spec_dict = self.config[component_name]
+                component_spec_dict["pretrained_model_name_or_path"] = repo_id
+                component_spec_dict["subfolder"] = component_name
+                if variant is not None and "variant" in component_spec_dict:
+                    component_spec_dict["variant"] = variant
+
+                self.register_to_config(**{component_name: (library, class_name, component_spec_dict)})
+
+        self.save_config(save_directory=save_directory)
+
+        if push_to_hub:
+            card_content = generate_modular_model_card_content(self.blocks)
             model_card = load_or_create_model_card(
                 repo_id,
                 token=token,
@@ -1850,13 +1939,8 @@ class ModularPipeline(ConfigMixin, PushToHubMixin):
                 is_modular=True,
             )
             model_card = populate_model_card(model_card, tags=card_content["tags"])
-
             model_card.save(os.path.join(save_directory, "README.md"))
 
-        # YiYi TODO: maybe order the json file to make it more readable: configs first, then components
-        self.save_config(save_directory=save_directory)
-
-        if push_to_hub:
             self._upload_folder(
                 save_directory,
                 repo_id,

src/diffusers/pipelines/audioldm2/pipeline_audioldm2.py

@@ -502,10 +502,6 @@ class AudioLDM2Pipeline(DiffusionPipeline):
                         text_input_ids,
                         attention_mask=attention_mask,
                     )
-                    # Extract the pooler output if it's a BaseModelOutputWithPooling (Transformers v5+)
-                    # otherwise use it directly (Transformers v4)
-                    if hasattr(prompt_embeds, "pooler_output"):
-                        prompt_embeds = prompt_embeds.pooler_output
                     # append the seq-len dim: (bs, hidden_size) -> (bs, seq_len, hidden_size)
                     prompt_embeds = prompt_embeds[:, None, :]
                     # make sure that we attend to this single hidden-state
@@ -614,10 +610,6 @@ class AudioLDM2Pipeline(DiffusionPipeline):
                         uncond_input_ids,
                         attention_mask=negative_attention_mask,
                     )
-                    # Extract the pooler output if it's a BaseModelOutputWithPooling (Transformers v5+)
-                    # otherwise use it directly (Transformers v4)
-                    if hasattr(negative_prompt_embeds, "pooler_output"):
-                        negative_prompt_embeds = negative_prompt_embeds.pooler_output
                     # append the seq-len dim: (bs, hidden_size) -> (bs, seq_len, hidden_size)
                     negative_prompt_embeds = negative_prompt_embeds[:, None, :]
                     # make sure that we attend to this single hidden-state

src/diffusers/pipelines/cosmos/pipeline_cosmos2_5_predict.py

@@ -287,9 +287,6 @@ class Cosmos2_5_PredictBasePipeline(DiffusionPipeline):
                 truncation=True,
                 padding="max_length",
             )
-            input_ids = (
-                input_ids["input_ids"] if not isinstance(input_ids, list) and "input_ids" in input_ids else input_ids
-            )
             input_ids = torch.LongTensor(input_ids)
             input_ids_batch.append(input_ids)
 

src/diffusers/pipelines/cosmos/pipeline_cosmos2_5_transfer.py

@@ -262,9 +262,6 @@ class Cosmos2_5_TransferPipeline(DiffusionPipeline):
                 truncation=True,
                 padding="max_length",
             )
-            input_ids = (
-                input_ids["input_ids"] if not isinstance(input_ids, list) and "input_ids" in input_ids else input_ids
-            )
             input_ids = torch.LongTensor(input_ids)
             input_ids_batch.append(input_ids)
 

src/diffusers/pipelines/kandinsky/text_encoder.py

@@ -20,8 +20,6 @@ class MultilingualCLIP(PreTrainedModel):
         self.LinearTransformation = torch.nn.Linear(
             in_features=config.transformerDimensions, out_features=config.numDims
         )
-        if hasattr(self, "post_init"):
-            self.post_init()
 
     def forward(self, input_ids, attention_mask):
         embs = self.transformer(input_ids=input_ids, attention_mask=attention_mask)[0]

src/diffusers/pipelines/kolors/text_encoder.py

@@ -781,9 +781,6 @@ class ChatGLMModel(ChatGLMPreTrainedModel):
             self.prefix_encoder = PrefixEncoder(config)
             self.dropout = torch.nn.Dropout(0.1)
 
-        if hasattr(self, "post_init"):
-            self.post_init()
-
     def get_input_embeddings(self):
         return self.embedding.word_embeddings
 
@@ -813,7 +810,7 @@ class ChatGLMModel(ChatGLMPreTrainedModel):
         output_hidden_states = (
             output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
         )
-        use_cache = use_cache if use_cache is not None else getattr(self.config, "use_cache", None)
+        use_cache = use_cache if use_cache is not None else self.config.use_cache
         return_dict = return_dict if return_dict is not None else self.config.use_return_dict
 
         batch_size, seq_length = input_ids.shape

src/diffusers/pipelines/pipeline_utils.py

@@ -341,7 +341,6 @@ class DiffusionPipeline(ConfigMixin, PushToHubMixin):
             save_method_accept_safe = "safe_serialization" in save_method_signature.parameters
             save_method_accept_variant = "variant" in save_method_signature.parameters
             save_method_accept_max_shard_size = "max_shard_size" in save_method_signature.parameters
-            save_method_accept_peft_format = "save_peft_format" in save_method_signature.parameters
 
             save_kwargs = {}
             if save_method_accept_safe:
@@ -351,11 +350,6 @@ class DiffusionPipeline(ConfigMixin, PushToHubMixin):
             if save_method_accept_max_shard_size and max_shard_size is not None:
                 # max_shard_size is expected to not be None in ModelMixin
                 save_kwargs["max_shard_size"] = max_shard_size
-            if save_method_accept_peft_format:
-                # Set save_peft_format=False for transformers>=5.0.0 compatibility
-                # In transformers 5.0.0+, the default save_peft_format=True adds "base_model.model" prefix
-                # to adapter keys, but from_pretrained expects keys without this prefix
-                save_kwargs["save_peft_format"] = False
 
             save_method(os.path.join(save_directory, pipeline_component_name), **save_kwargs)
 

src/diffusers/pipelines/skyreels_v2/pipeline_skyreels_v2.py

@@ -17,7 +17,7 @@ from typing import Any, Callable
 
 import regex as re
 import torch
-from transformers import AutoTokenizer, T5EncoderModel, UMT5EncoderModel
+from transformers import AutoTokenizer, UMT5EncoderModel
 
 from ...callbacks import MultiPipelineCallbacks, PipelineCallback
 from ...loaders import SkyReelsV2LoraLoaderMixin
@@ -132,7 +132,7 @@ class SkyReelsV2Pipeline(DiffusionPipeline, SkyReelsV2LoraLoaderMixin):
     def __init__(
         self,
         tokenizer: AutoTokenizer,
-        text_encoder: T5EncoderModel | UMT5EncoderModel,
+        text_encoder: UMT5EncoderModel,
         transformer: SkyReelsV2Transformer3DModel,
         vae: AutoencoderKLWan,
         scheduler: UniPCMultistepScheduler,

src/diffusers/pipelines/skyreels_v2/pipeline_skyreels_v2_diffusion_forcing.py

@@ -19,7 +19,7 @@ from copy import deepcopy
 from typing import Any, Callable
 
 import torch
-from transformers import AutoTokenizer, T5EncoderModel, UMT5EncoderModel
+from transformers import AutoTokenizer, UMT5EncoderModel
 
 from ...callbacks import MultiPipelineCallbacks, PipelineCallback
 from ...loaders import SkyReelsV2LoraLoaderMixin
@@ -153,7 +153,7 @@ class SkyReelsV2DiffusionForcingPipeline(DiffusionPipeline, SkyReelsV2LoraLoader
     def __init__(
         self,
         tokenizer: AutoTokenizer,
-        text_encoder: T5EncoderModel | UMT5EncoderModel,
+        text_encoder: UMT5EncoderModel,
         transformer: SkyReelsV2Transformer3DModel,
         vae: AutoencoderKLWan,
         scheduler: UniPCMultistepScheduler,

src/diffusers/pipelines/skyreels_v2/pipeline_skyreels_v2_diffusion_forcing_i2v.py

@@ -20,7 +20,7 @@ from typing import Any, Callable
 
 import PIL
 import torch
-from transformers import AutoTokenizer, T5EncoderModel, UMT5EncoderModel
+from transformers import AutoTokenizer, UMT5EncoderModel
 
 from diffusers.image_processor import PipelineImageInput
 from diffusers.utils.torch_utils import randn_tensor
@@ -158,7 +158,7 @@ class SkyReelsV2DiffusionForcingImageToVideoPipeline(DiffusionPipeline, SkyReels
     def __init__(
         self,
         tokenizer: AutoTokenizer,
-        text_encoder: T5EncoderModel | UMT5EncoderModel,
+        text_encoder: UMT5EncoderModel,
         transformer: SkyReelsV2Transformer3DModel,
         vae: AutoencoderKLWan,
         scheduler: UniPCMultistepScheduler,

src/diffusers/pipelines/skyreels_v2/pipeline_skyreels_v2_diffusion_forcing_v2v.py

@@ -21,7 +21,7 @@ from typing import Any, Callable
 
 import torch
 from PIL import Image
-from transformers import AutoTokenizer, T5EncoderModel, UMT5EncoderModel
+from transformers import AutoTokenizer, UMT5EncoderModel
 
 from ...callbacks import MultiPipelineCallbacks, PipelineCallback
 from ...loaders import SkyReelsV2LoraLoaderMixin
@@ -214,7 +214,7 @@ class SkyReelsV2DiffusionForcingVideoToVideoPipeline(DiffusionPipeline, SkyReels
     def __init__(
         self,
         tokenizer: AutoTokenizer,
-        text_encoder: T5EncoderModel | UMT5EncoderModel,
+        text_encoder: UMT5EncoderModel,
         transformer: SkyReelsV2Transformer3DModel,
         vae: AutoencoderKLWan,
         scheduler: UniPCMultistepScheduler,

src/diffusers/pipelines/skyreels_v2/pipeline_skyreels_v2_i2v.py

@@ -18,7 +18,7 @@ from typing import Any, Callable
 import PIL
 import regex as re
 import torch
-from transformers import AutoTokenizer, CLIPProcessor, CLIPVisionModelWithProjection, T5EncoderModel, UMT5EncoderModel
+from transformers import AutoTokenizer, CLIPProcessor, CLIPVisionModelWithProjection, UMT5EncoderModel
 
 from ...callbacks import MultiPipelineCallbacks, PipelineCallback
 from ...image_processor import PipelineImageInput
@@ -157,7 +157,7 @@ class SkyReelsV2ImageToVideoPipeline(DiffusionPipeline, SkyReelsV2LoraLoaderMixi
     def __init__(
         self,
         tokenizer: AutoTokenizer,
-        text_encoder: T5EncoderModel | UMT5EncoderModel,
+        text_encoder: UMT5EncoderModel,
         image_encoder: CLIPVisionModelWithProjection,
         image_processor: CLIPProcessor,
         transformer: SkyReelsV2Transformer3DModel,

src/diffusers/pipelines/transformers_loading_utils.py

@@ -112,8 +112,6 @@ def _load_transformers_model_from_dduf(
                 tensors = safetensors.torch.load(mmap)
                 # Update the state dictionary with tensors
                 state_dict.update(tensors)
-            # `from_pretrained` sets the model to eval mode by default, which is the
-            # correct behavior for inference. Do not call `model.train()` here.
             return cls.from_pretrained(
                 pretrained_model_name_or_path=None,
                 config=config,

tests/models/test_models_auto.py

@@ -20,9 +20,7 @@ class TestAutoModel(unittest.TestCase):
         side_effect=[EnvironmentError("File not found"), {"model_type": "clip_text_model"}],
     )
     def test_load_from_config_transformers_with_subfolder(self, mock_load_config):
-        model = AutoModel.from_pretrained(
-            "hf-internal-testing/tiny-stable-diffusion-torch", subfolder="text_encoder", use_safetensors=False
-        )
+        model = AutoModel.from_pretrained("hf-internal-testing/tiny-stable-diffusion-torch", subfolder="text_encoder")
         assert isinstance(model, CLIPTextModel)
 
     def test_load_from_config_without_subfolder(self):
@@ -30,7 +28,5 @@ class TestAutoModel(unittest.TestCase):
         assert isinstance(model, LongformerModel)
 
     def test_load_from_model_index(self):
-        model = AutoModel.from_pretrained(
-            "hf-internal-testing/tiny-stable-diffusion-torch", subfolder="text_encoder", use_safetensors=False
-        )
+        model = AutoModel.from_pretrained("hf-internal-testing/tiny-stable-diffusion-torch", subfolder="text_encoder")
         assert isinstance(model, CLIPTextModel)

tests/pipelines/bria/test_pipeline_bria.py

@@ -19,7 +19,7 @@ import unittest
 import numpy as np
 import torch
 from huggingface_hub import hf_hub_download
-from transformers import AutoConfig, T5EncoderModel, T5TokenizerFast
+from transformers import T5EncoderModel, T5TokenizerFast
 
 from diffusers import (
     AutoencoderKL,
@@ -89,8 +89,7 @@ class BriaPipelineFastTests(PipelineTesterMixin, unittest.TestCase):
         scheduler = FlowMatchEulerDiscreteScheduler()
 
         torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer = T5TokenizerFast.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         components = {

tests/pipelines/chroma/test_pipeline_chroma.py

@@ -2,7 +2,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import AutoConfig, AutoTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, T5EncoderModel
 
 from diffusers import AutoencoderKL, ChromaPipeline, ChromaTransformer2DModel, FlowMatchEulerDiscreteScheduler
 
@@ -41,8 +41,7 @@ class ChromaPipelineFastTests(
         )
 
         torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 

tests/pipelines/chroma/test_pipeline_chroma_img2img.py

@@ -3,7 +3,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import AutoConfig, AutoTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, T5EncoderModel
 
 from diffusers import AutoencoderKL, ChromaImg2ImgPipeline, ChromaTransformer2DModel, FlowMatchEulerDiscreteScheduler
 
@@ -42,8 +42,7 @@ class ChromaImg2ImgPipelineFastTests(
         )
 
         torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 

tests/pipelines/chronoedit/test_chronoedit.py

@@ -17,7 +17,6 @@ import unittest
 import torch
 from PIL import Image
 from transformers import (
-    AutoConfig,
     AutoTokenizer,
     CLIPImageProcessor,
     CLIPVisionConfig,
@@ -72,8 +71,7 @@ class ChronoEditPipelineFastTests(PipelineTesterMixin, unittest.TestCase):
         torch.manual_seed(0)
         # TODO: impl FlowDPMSolverMultistepScheduler
         scheduler = FlowMatchEulerDiscreteScheduler(shift=7.0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         torch.manual_seed(0)

tests/pipelines/cogvideo/test_cogvideox.py

@@ -18,7 +18,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import AutoConfig, AutoTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, T5EncoderModel
 
 from diffusers import AutoencoderKLCogVideoX, CogVideoXPipeline, CogVideoXTransformer3DModel, DDIMScheduler
 
@@ -117,8 +117,7 @@ class CogVideoXPipelineFastTests(
 
         torch.manual_seed(0)
         scheduler = DDIMScheduler()
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         components = {

tests/pipelines/cogvideo/test_cogvideox_fun_control.py

@@ -18,7 +18,7 @@ import unittest
 import numpy as np
 import torch
 from PIL import Image
-from transformers import AutoConfig, AutoTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, T5EncoderModel
 
 from diffusers import AutoencoderKLCogVideoX, CogVideoXFunControlPipeline, CogVideoXTransformer3DModel, DDIMScheduler
 
@@ -104,8 +104,7 @@ class CogVideoXFunControlPipelineFastTests(PipelineTesterMixin, unittest.TestCas
 
         torch.manual_seed(0)
         scheduler = DDIMScheduler()
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         components = {

tests/pipelines/cogvideo/test_cogvideox_image2video.py

@@ -19,7 +19,7 @@ import unittest
 import numpy as np
 import torch
 from PIL import Image
-from transformers import AutoConfig, AutoTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, T5EncoderModel
 
 from diffusers import AutoencoderKLCogVideoX, CogVideoXImageToVideoPipeline, CogVideoXTransformer3DModel, DDIMScheduler
 from diffusers.utils import load_image
@@ -113,8 +113,7 @@ class CogVideoXImageToVideoPipelineFastTests(PipelineTesterMixin, unittest.TestC
 
         torch.manual_seed(0)
         scheduler = DDIMScheduler()
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         components = {
@@ -238,9 +237,6 @@ class CogVideoXImageToVideoPipelineFastTests(PipelineTesterMixin, unittest.TestC
             return
 
         components = self.get_dummy_components()
-        for key in components:
-            if "text_encoder" in key and hasattr(components[key], "eval"):
-                components[key].eval()
         pipe = self.pipeline_class(**components)
         for component in pipe.components.values():
             if hasattr(component, "set_default_attn_processor"):

tests/pipelines/cogvideo/test_cogvideox_video2video.py

@@ -18,7 +18,7 @@ import unittest
 import numpy as np
 import torch
 from PIL import Image
-from transformers import AutoConfig, AutoTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, T5EncoderModel
 
 from diffusers import AutoencoderKLCogVideoX, CogVideoXTransformer3DModel, CogVideoXVideoToVideoPipeline, DDIMScheduler
 
@@ -99,8 +99,7 @@ class CogVideoXVideoToVideoPipelineFastTests(PipelineTesterMixin, unittest.TestC
 
         torch.manual_seed(0)
         scheduler = DDIMScheduler()
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         components = {

tests/pipelines/cogview3/test_cogview3plus.py

@@ -18,7 +18,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import AutoConfig, AutoTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, T5EncoderModel
 
 from diffusers import AutoencoderKL, CogVideoXDDIMScheduler, CogView3PlusPipeline, CogView3PlusTransformer2DModel
 
@@ -89,8 +89,7 @@ class CogView3PlusPipelineFastTests(PipelineTesterMixin, unittest.TestCase):
 
         torch.manual_seed(0)
         scheduler = CogVideoXDDIMScheduler()
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         components = {

tests/pipelines/cogview4/test_cogview4.py

@@ -108,7 +108,7 @@ class CogView4PipelineFastTests(PipelineTesterMixin, unittest.TestCase):
             generator = torch.Generator(device=device).manual_seed(seed)
         inputs = {
             "prompt": "dance monkey",
-            "negative_prompt": "bad",
+            "negative_prompt": "",
             "generator": generator,
             "num_inference_steps": 2,
             "guidance_scale": 6.0,

tests/pipelines/consisid/test_consisid.py

@@ -19,7 +19,7 @@ import unittest
 import numpy as np
 import torch
 from PIL import Image
-from transformers import AutoConfig, AutoTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, T5EncoderModel
 
 from diffusers import AutoencoderKLCogVideoX, ConsisIDPipeline, ConsisIDTransformer3DModel, DDIMScheduler
 from diffusers.utils import load_image
@@ -122,8 +122,7 @@ class ConsisIDPipelineFastTests(PipelineTesterMixin, unittest.TestCase):
 
         torch.manual_seed(0)
         scheduler = DDIMScheduler()
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         components = {
@@ -249,9 +248,6 @@ class ConsisIDPipelineFastTests(PipelineTesterMixin, unittest.TestCase):
             return
 
         components = self.get_dummy_components()
-        for key in components:
-            if "text_encoder" in key and hasattr(components[key], "eval"):
-                components[key].eval()
         pipe = self.pipeline_class(**components)
         for component in pipe.components.values():
             if hasattr(component, "set_default_attn_processor"):

tests/pipelines/controlnet_flux/test_controlnet_flux.py

@@ -19,7 +19,7 @@ import unittest
 import numpy as np
 import torch
 from huggingface_hub import hf_hub_download
-from transformers import AutoConfig, CLIPTextConfig, CLIPTextModel, CLIPTokenizer, T5EncoderModel, T5TokenizerFast
+from transformers import CLIPTextConfig, CLIPTextModel, CLIPTokenizer, T5EncoderModel, T5TokenizerFast
 
 from diffusers import (
     AutoencoderKL,
@@ -97,8 +97,7 @@ class FluxControlNetPipelineFastTests(unittest.TestCase, PipelineTesterMixin, Fl
         text_encoder = CLIPTextModel(clip_text_encoder_config)
 
         torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder_2 = T5EncoderModel(config)
+        text_encoder_2 = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         tokenizer = CLIPTokenizer.from_pretrained("hf-internal-testing/tiny-random-clip")
         tokenizer_2 = T5TokenizerFast.from_pretrained("hf-internal-testing/tiny-random-t5")

tests/pipelines/controlnet_flux/test_controlnet_flux_img2img.py

@@ -2,7 +2,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import AutoConfig, AutoTokenizer, CLIPTextConfig, CLIPTextModel, CLIPTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, CLIPTextConfig, CLIPTextModel, CLIPTokenizer, T5EncoderModel
 
 from diffusers import (
     AutoencoderKL,
@@ -13,7 +13,9 @@ from diffusers import (
 )
 from diffusers.utils.torch_utils import randn_tensor
 
-from ...testing_utils import torch_device
+from ...testing_utils import (
+    torch_device,
+)
 from ..test_pipelines_common import PipelineTesterMixin, check_qkv_fused_layers_exist
 
 
@@ -68,8 +70,7 @@ class FluxControlNetImg2ImgPipelineFastTests(unittest.TestCase, PipelineTesterMi
         text_encoder = CLIPTextModel(clip_text_encoder_config)
 
         torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder_2 = T5EncoderModel(config)
+        text_encoder_2 = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         tokenizer = CLIPTokenizer.from_pretrained("hf-internal-testing/tiny-random-clip")
         tokenizer_2 = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")

tests/pipelines/controlnet_flux/test_controlnet_flux_inpaint.py

@@ -3,7 +3,15 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import AutoConfig, AutoTokenizer, CLIPTextConfig, CLIPTextModel, CLIPTokenizer, T5EncoderModel
+
+# torch_device,  # {{ edit_1 }} Removed unused import
+from transformers import (
+    AutoTokenizer,
+    CLIPTextConfig,
+    CLIPTextModel,
+    CLIPTokenizer,
+    T5EncoderModel,
+)
 
 from diffusers import (
     AutoencoderKL,
@@ -14,7 +22,11 @@ from diffusers import (
 )
 from diffusers.utils.torch_utils import randn_tensor
 
-from ...testing_utils import enable_full_determinism, floats_tensor, torch_device
+from ...testing_utils import (
+    enable_full_determinism,
+    floats_tensor,
+    torch_device,
+)
 from ..test_pipelines_common import PipelineTesterMixin
 
 
@@ -73,8 +85,7 @@ class FluxControlNetInpaintPipelineTests(unittest.TestCase, PipelineTesterMixin)
         text_encoder = CLIPTextModel(clip_text_encoder_config)
 
         torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder_2 = T5EncoderModel(config)
+        text_encoder_2 = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         tokenizer = CLIPTokenizer.from_pretrained("hf-internal-testing/tiny-random-clip")
         tokenizer_2 = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")

tests/pipelines/controlnet_hunyuandit/test_controlnet_hunyuandit.py

@@ -18,7 +18,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import AutoConfig, AutoTokenizer, BertModel, T5EncoderModel
+from transformers import AutoTokenizer, BertModel, T5EncoderModel
 
 from diffusers import (
     AutoencoderKL,
@@ -96,10 +96,7 @@ class HunyuanDiTControlNetPipelineFastTests(unittest.TestCase, PipelineTesterMix
         scheduler = DDPMScheduler()
         text_encoder = BertModel.from_pretrained("hf-internal-testing/tiny-random-BertModel")
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-BertModel")
-
-        torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder_2 = T5EncoderModel(config)
+        text_encoder_2 = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer_2 = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         components = {

tests/pipelines/controlnet_sd3/test_controlnet_inpaint_sd3.py

@@ -17,14 +17,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import (
-    AutoConfig,
-    AutoTokenizer,
-    CLIPTextConfig,
-    CLIPTextModelWithProjection,
-    CLIPTokenizer,
-    T5EncoderModel,
-)
+from transformers import AutoTokenizer, CLIPTextConfig, CLIPTextModelWithProjection, CLIPTokenizer, T5EncoderModel
 
 from diffusers import (
     AutoencoderKL,
@@ -35,7 +28,10 @@ from diffusers import (
 from diffusers.models import SD3ControlNetModel
 from diffusers.utils.torch_utils import randn_tensor
 
-from ...testing_utils import enable_full_determinism, torch_device
+from ...testing_utils import (
+    enable_full_determinism,
+    torch_device,
+)
 from ..test_pipelines_common import PipelineTesterMixin
 
 
@@ -107,8 +103,7 @@ class StableDiffusion3ControlInpaintNetPipelineFastTests(unittest.TestCase, Pipe
         text_encoder_2 = CLIPTextModelWithProjection(clip_text_encoder_config)
 
         torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder_3 = T5EncoderModel(config)
+        text_encoder_3 = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         tokenizer = CLIPTokenizer.from_pretrained("hf-internal-testing/tiny-random-clip")
         tokenizer_2 = CLIPTokenizer.from_pretrained("hf-internal-testing/tiny-random-clip")

tests/pipelines/controlnet_sd3/test_controlnet_sd3.py

@@ -18,14 +18,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import (
-    AutoConfig,
-    AutoTokenizer,
-    CLIPTextConfig,
-    CLIPTextModelWithProjection,
-    CLIPTokenizer,
-    T5EncoderModel,
-)
+from transformers import AutoTokenizer, CLIPTextConfig, CLIPTextModelWithProjection, CLIPTokenizer, T5EncoderModel
 
 from diffusers import (
     AutoencoderKL,
@@ -124,8 +117,7 @@ class StableDiffusion3ControlNetPipelineFastTests(unittest.TestCase, PipelineTes
         text_encoder_2 = CLIPTextModelWithProjection(clip_text_encoder_config)
 
         torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder_3 = T5EncoderModel(config)
+        text_encoder_3 = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         tokenizer = CLIPTokenizer.from_pretrained("hf-internal-testing/tiny-random-clip")
         tokenizer_2 = CLIPTokenizer.from_pretrained("hf-internal-testing/tiny-random-clip")

tests/pipelines/cosmos/test_cosmos.py

@@ -20,7 +20,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import AutoConfig, AutoTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, T5EncoderModel
 
 from diffusers import AutoencoderKLCosmos, CosmosTextToWorldPipeline, CosmosTransformer3DModel, EDMEulerScheduler
 
@@ -107,8 +107,7 @@ class CosmosTextToWorldPipelineFastTests(PipelineTesterMixin, unittest.TestCase)
             rho=7.0,
             final_sigmas_type="sigma_min",
         )
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         components = {
@@ -233,9 +232,6 @@ class CosmosTextToWorldPipelineFastTests(PipelineTesterMixin, unittest.TestCase)
             return
 
         components = self.get_dummy_components()
-        for key in components:
-            if "text_encoder" in key and hasattr(components[key], "eval"):
-                components[key].eval()
         pipe = self.pipeline_class(**components)
         for component in pipe.components.values():
             if hasattr(component, "set_default_attn_processor"):

tests/pipelines/cosmos/test_cosmos2_text2image.py

@@ -20,7 +20,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import AutoConfig, AutoTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, T5EncoderModel
 
 from diffusers import (
     AutoencoderKLWan,
@@ -95,8 +95,7 @@ class Cosmos2TextToImagePipelineFastTests(PipelineTesterMixin, unittest.TestCase
 
         torch.manual_seed(0)
         scheduler = FlowMatchEulerDiscreteScheduler(use_karras_sigmas=True)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         components = {

tests/pipelines/cosmos/test_cosmos2_video2world.py

@@ -21,7 +21,7 @@ import unittest
 import numpy as np
 import PIL.Image
 import torch
-from transformers import AutoConfig, AutoTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, T5EncoderModel
 
 from diffusers import (
     AutoencoderKLWan,
@@ -96,8 +96,7 @@ class Cosmos2VideoToWorldPipelineFastTests(PipelineTesterMixin, unittest.TestCas
 
         torch.manual_seed(0)
         scheduler = FlowMatchEulerDiscreteScheduler(use_karras_sigmas=True)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         components = {

tests/pipelines/cosmos/test_cosmos_video2world.py

@@ -21,7 +21,7 @@ import unittest
 import numpy as np
 import PIL.Image
 import torch
-from transformers import AutoConfig, AutoTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, T5EncoderModel
 
 from diffusers import AutoencoderKLCosmos, CosmosTransformer3DModel, CosmosVideoToWorldPipeline, EDMEulerScheduler
 
@@ -108,8 +108,7 @@ class CosmosVideoToWorldPipelineFastTests(PipelineTesterMixin, unittest.TestCase
             rho=7.0,
             final_sigmas_type="sigma_min",
         )
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         components = {
@@ -246,9 +245,6 @@ class CosmosVideoToWorldPipelineFastTests(PipelineTesterMixin, unittest.TestCase
             return
 
         components = self.get_dummy_components()
-        for key in components:
-            if "text_encoder" in key and hasattr(components[key], "eval"):
-                components[key].eval()
         pipe = self.pipeline_class(**components)
         for component in pipe.components.values():
             if hasattr(component, "set_default_attn_processor"):

tests/pipelines/deepfloyd_if/__init__.py

@@ -2,7 +2,7 @@ import tempfile
 
 import numpy as np
 import torch
-from transformers import AutoConfig, AutoTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, T5EncoderModel
 
 from diffusers import DDPMScheduler, UNet2DConditionModel
 from diffusers.models.attention_processor import AttnAddedKVProcessor
@@ -18,8 +18,7 @@ from ..test_pipelines_common import to_np
 class IFPipelineTesterMixin:
     def _get_dummy_components(self):
         torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         torch.manual_seed(0)
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
@@ -76,8 +75,7 @@ class IFPipelineTesterMixin:
 
     def _get_superresolution_dummy_components(self):
         torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         torch.manual_seed(0)
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
@@ -252,9 +250,6 @@ class IFPipelineTesterMixin:
     # This should be handled in the base test and then this method can be removed.
     def _test_save_load_local(self):
         components = self.get_dummy_components()
-        for key in components:
-            if "text_encoder" in key and hasattr(components[key], "eval"):
-                components[key].eval()
         pipe = self.pipeline_class(**components)
         pipe.to(torch_device)
         pipe.set_progress_bar_config(disable=None)

tests/pipelines/deepfloyd_if/test_if.py

@@ -18,7 +18,9 @@ import unittest
 
 import torch
 
-from diffusers import IFPipeline
+from diffusers import (
+    IFPipeline,
+)
 from diffusers.models.attention_processor import AttnAddedKVProcessor
 from diffusers.utils.import_utils import is_xformers_available
 

tests/pipelines/flux/test_pipeline_flux.py

@@ -4,7 +4,7 @@ import unittest
 import numpy as np
 import torch
 from huggingface_hub import hf_hub_download
-from transformers import AutoConfig, AutoTokenizer, CLIPTextConfig, CLIPTextModel, CLIPTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, CLIPTextConfig, CLIPTextModel, CLIPTokenizer, T5EncoderModel
 
 from diffusers import (
     AutoencoderKL,
@@ -93,8 +93,7 @@ class FluxPipelineFastTests(
         text_encoder = CLIPTextModel(clip_text_encoder_config)
 
         torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder_2 = T5EncoderModel(config)
+        text_encoder_2 = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         tokenizer = CLIPTokenizer.from_pretrained("hf-internal-testing/tiny-random-clip")
         tokenizer_2 = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")

tests/pipelines/flux/test_pipeline_flux_control.py

@@ -3,7 +3,7 @@ import unittest
 import numpy as np
 import torch
 from PIL import Image
-from transformers import AutoConfig, AutoTokenizer, CLIPTextConfig, CLIPTextModel, CLIPTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, CLIPTextConfig, CLIPTextModel, CLIPTokenizer, T5EncoderModel
 
 from diffusers import AutoencoderKL, FlowMatchEulerDiscreteScheduler, FluxControlPipeline, FluxTransformer2DModel
 
@@ -53,8 +53,7 @@ class FluxControlPipelineFastTests(unittest.TestCase, PipelineTesterMixin):
         text_encoder = CLIPTextModel(clip_text_encoder_config)
 
         torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder_2 = T5EncoderModel(config)
+        text_encoder_2 = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         tokenizer = CLIPTokenizer.from_pretrained("hf-internal-testing/tiny-random-clip")
         tokenizer_2 = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")

tests/pipelines/flux/test_pipeline_flux_control_img2img.py

@@ -3,7 +3,7 @@ import unittest
 import numpy as np
 import torch
 from PIL import Image
-from transformers import AutoConfig, AutoTokenizer, CLIPTextConfig, CLIPTextModel, CLIPTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, CLIPTextConfig, CLIPTextModel, CLIPTokenizer, T5EncoderModel
 
 from diffusers import (
     AutoencoderKL,
@@ -57,8 +57,7 @@ class FluxControlImg2ImgPipelineFastTests(unittest.TestCase, PipelineTesterMixin
         text_encoder = CLIPTextModel(clip_text_encoder_config)
 
         torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder_2 = T5EncoderModel(config)
+        text_encoder_2 = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         tokenizer = CLIPTokenizer.from_pretrained("hf-internal-testing/tiny-random-clip")
         tokenizer_2 = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")

tests/pipelines/flux/test_pipeline_flux_control_inpaint.py

@@ -3,7 +3,7 @@ import unittest
 import numpy as np
 import torch
 from PIL import Image
-from transformers import AutoConfig, AutoTokenizer, CLIPTextConfig, CLIPTextModel, CLIPTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, CLIPTextConfig, CLIPTextModel, CLIPTokenizer, T5EncoderModel
 
 from diffusers import (
     AutoencoderKL,
@@ -58,8 +58,7 @@ class FluxControlInpaintPipelineFastTests(unittest.TestCase, PipelineTesterMixin
         text_encoder = CLIPTextModel(clip_text_encoder_config)
 
         torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder_2 = T5EncoderModel(config)
+        text_encoder_2 = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         tokenizer = CLIPTokenizer.from_pretrained("hf-internal-testing/tiny-random-clip")
         tokenizer_2 = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")

tests/pipelines/flux/test_pipeline_flux_fill.py

@@ -3,7 +3,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import AutoConfig, AutoTokenizer, CLIPTextConfig, CLIPTextModel, CLIPTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, CLIPTextConfig, CLIPTextModel, CLIPTokenizer, T5EncoderModel
 
 from diffusers import AutoencoderKL, FlowMatchEulerDiscreteScheduler, FluxFillPipeline, FluxTransformer2DModel
 
@@ -58,8 +58,7 @@ class FluxFillPipelineFastTests(unittest.TestCase, PipelineTesterMixin):
         text_encoder = CLIPTextModel(clip_text_encoder_config)
 
         torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder_2 = T5EncoderModel(config)
+        text_encoder_2 = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         tokenizer = CLIPTokenizer.from_pretrained("hf-internal-testing/tiny-random-clip")
         tokenizer_2 = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")

tests/pipelines/flux/test_pipeline_flux_img2img.py

@@ -3,7 +3,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import AutoConfig, AutoTokenizer, CLIPTextConfig, CLIPTextModel, CLIPTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, CLIPTextConfig, CLIPTextModel, CLIPTokenizer, T5EncoderModel
 
 from diffusers import AutoencoderKL, FlowMatchEulerDiscreteScheduler, FluxImg2ImgPipeline, FluxTransformer2DModel
 
@@ -55,8 +55,7 @@ class FluxImg2ImgPipelineFastTests(unittest.TestCase, PipelineTesterMixin, FluxI
         text_encoder = CLIPTextModel(clip_text_encoder_config)
 
         torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder_2 = T5EncoderModel(config)
+        text_encoder_2 = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         tokenizer = CLIPTokenizer.from_pretrained("hf-internal-testing/tiny-random-clip")
         tokenizer_2 = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")

tests/pipelines/flux/test_pipeline_flux_inpaint.py

@@ -3,7 +3,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import AutoConfig, AutoTokenizer, CLIPTextConfig, CLIPTextModel, CLIPTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, CLIPTextConfig, CLIPTextModel, CLIPTokenizer, T5EncoderModel
 
 from diffusers import AutoencoderKL, FlowMatchEulerDiscreteScheduler, FluxInpaintPipeline, FluxTransformer2DModel
 
@@ -55,8 +55,7 @@ class FluxInpaintPipelineFastTests(unittest.TestCase, PipelineTesterMixin, FluxI
         text_encoder = CLIPTextModel(clip_text_encoder_config)
 
         torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder_2 = T5EncoderModel(config)
+        text_encoder_2 = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         tokenizer = CLIPTokenizer.from_pretrained("hf-internal-testing/tiny-random-clip")
         tokenizer_2 = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")

tests/pipelines/flux/test_pipeline_flux_kontext.py

@@ -3,7 +3,7 @@ import unittest
 import numpy as np
 import PIL.Image
 import torch
-from transformers import AutoConfig, AutoTokenizer, CLIPTextConfig, CLIPTextModel, CLIPTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, CLIPTextConfig, CLIPTextModel, CLIPTokenizer, T5EncoderModel
 
 from diffusers import (
     AutoencoderKL,
@@ -79,8 +79,7 @@ class FluxKontextPipelineFastTests(
         text_encoder = CLIPTextModel(clip_text_encoder_config)
 
         torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder_2 = T5EncoderModel(config)
+        text_encoder_2 = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         tokenizer = CLIPTokenizer.from_pretrained("hf-internal-testing/tiny-random-clip")
         tokenizer_2 = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")

tests/pipelines/flux/test_pipeline_flux_kontext_inpaint.py

@@ -3,7 +3,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import AutoConfig, AutoTokenizer, CLIPTextConfig, CLIPTextModel, CLIPTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, CLIPTextConfig, CLIPTextModel, CLIPTokenizer, T5EncoderModel
 
 from diffusers import (
     AutoencoderKL,
@@ -79,8 +79,7 @@ class FluxKontextInpaintPipelineFastTests(
         text_encoder = CLIPTextModel(clip_text_encoder_config)
 
         torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder_2 = T5EncoderModel(config)
+        text_encoder_2 = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         tokenizer = CLIPTokenizer.from_pretrained("hf-internal-testing/tiny-random-clip")
         tokenizer_2 = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")

tests/pipelines/glm_image/test_glm_image.py

@@ -16,7 +16,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import AutoConfig, AutoTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, T5EncoderModel
 
 from diffusers import AutoencoderKL, FlowMatchEulerDiscreteScheduler, GlmImagePipeline, GlmImageTransformer2DModel
 from diffusers.utils import is_transformers_version
@@ -57,8 +57,7 @@ class GlmImagePipelineFastTests(PipelineTesterMixin, unittest.TestCase):
 
     def get_dummy_components(self):
         torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         glm_config = GlmImageConfig(

tests/pipelines/hidream_image/test_pipeline_hidream.py

@@ -18,7 +18,6 @@ import unittest
 import numpy as np
 import torch
 from transformers import (
-    AutoConfig,
     AutoTokenizer,
     CLIPTextConfig,
     CLIPTextModelWithProjection,
@@ -95,8 +94,7 @@ class HiDreamImagePipelineFastTests(PipelineTesterMixin, unittest.TestCase):
         text_encoder_2 = CLIPTextModelWithProjection(clip_text_encoder_config)
 
         torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder_3 = T5EncoderModel(config)
+        text_encoder_3 = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         torch.manual_seed(0)
         text_encoder_4 = LlamaForCausalLM.from_pretrained("hf-internal-testing/tiny-random-LlamaForCausalLM")
@@ -151,7 +149,7 @@ class HiDreamImagePipelineFastTests(PipelineTesterMixin, unittest.TestCase):
         self.assertEqual(generated_image.shape, (128, 128, 3))
 
         # fmt: off
-        expected_slice = np.array([0.4501, 0.5256, 0.4207, 0.5783, 0.4842, 0.4833, 0.4441, 0.5112, 0.6587, 0.3169, 0.7308, 0.5927, 0.6251, 0.5509, 0.5355, 0.5969])
+        expected_slice = np.array([0.4507, 0.5256, 0.4205, 0.5791, 0.4848, 0.4831, 0.4443, 0.5107, 0.6586, 0.3163, 0.7318, 0.5933, 0.6252, 0.5512, 0.5357, 0.5983])
         # fmt: on
 
         generated_slice = generated_image.flatten()

tests/pipelines/hunyuan_video/test_hunyuan_image2video.py

@@ -233,7 +233,7 @@ class HunyuanVideoImageToVideoPipelineFastTests(
         self.assertEqual(generated_video.shape, (5, 3, 16, 16))
 
         # fmt: off
-        expected_slice = torch.tensor([0.4441, 0.4790, 0.4485, 0.5748, 0.3539, 0.1553, 0.2707, 0.3594, 0.5331, 0.6645, 0.6799, 0.5257, 0.5092, 0.3450, 0.4276, 0.4127])
+        expected_slice = torch.tensor([0.444, 0.479, 0.4485, 0.5752, 0.3539, 0.1548, 0.2706, 0.3593, 0.5323, 0.6635, 0.6795, 0.5255, 0.5091, 0.345, 0.4276, 0.4128])
         # fmt: on
 
         generated_slice = generated_video.flatten()

tests/pipelines/hunyuan_video1_5/test_hunyuan_1_5.py

@@ -15,14 +15,7 @@
 import unittest
 
 import torch
-from transformers import (
-    AutoConfig,
-    ByT5Tokenizer,
-    Qwen2_5_VLTextConfig,
-    Qwen2_5_VLTextModel,
-    Qwen2Tokenizer,
-    T5EncoderModel,
-)
+from transformers import ByT5Tokenizer, Qwen2_5_VLTextConfig, Qwen2_5_VLTextModel, Qwen2Tokenizer, T5EncoderModel
 
 from diffusers import (
     AutoencoderKLHunyuanVideo15,
@@ -121,8 +114,7 @@ class HunyuanVideo15PipelineFastTests(PipelineTesterMixin, unittest.TestCase):
         tokenizer = Qwen2Tokenizer.from_pretrained("hf-internal-testing/tiny-random-Qwen2VLForConditionalGeneration")
 
         torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder_2 = T5EncoderModel(config)
+        text_encoder_2 = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer_2 = ByT5Tokenizer()
 
         guider = ClassifierFreeGuidance(guidance_scale=1.0)

tests/pipelines/hunyuandit/test_hunyuan_dit.py

@@ -19,7 +19,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import AutoConfig, AutoTokenizer, BertModel, T5EncoderModel
+from transformers import AutoTokenizer, BertModel, T5EncoderModel
 
 from diffusers import AutoencoderKL, DDPMScheduler, HunyuanDiT2DModel, HunyuanDiTPipeline
 
@@ -74,9 +74,7 @@ class HunyuanDiTPipelineFastTests(PipelineTesterMixin, unittest.TestCase):
         scheduler = DDPMScheduler()
         text_encoder = BertModel.from_pretrained("hf-internal-testing/tiny-random-BertModel")
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-BertModel")
-        torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder_2 = T5EncoderModel(config)
+        text_encoder_2 = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer_2 = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         components = {

tests/pipelines/kandinsky3/test_kandinsky3.py

@@ -19,7 +19,7 @@ import unittest
 import numpy as np
 import torch
 from PIL import Image
-from transformers import AutoConfig, AutoTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, T5EncoderModel
 
 from diffusers import (
     AutoPipelineForImage2Image,
@@ -108,8 +108,7 @@ class Kandinsky3PipelineFastTests(PipelineTesterMixin, unittest.TestCase):
         torch.manual_seed(0)
         movq = self.dummy_movq
         torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config).eval()
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         torch.manual_seed(0)
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
@@ -156,7 +155,7 @@ class Kandinsky3PipelineFastTests(PipelineTesterMixin, unittest.TestCase):
 
         assert image.shape == (1, 16, 16, 3)
 
-        expected_slice = np.array([0.3301, 0.3106, 0.4827, 0.5204, 0.4242, 0.4987, 0.4665, 0.5120, 0.5558])
+        expected_slice = np.array([0.3768, 0.4373, 0.4865, 0.4890, 0.4299, 0.5122, 0.4921, 0.4924, 0.5599])
 
         assert np.abs(image_slice.flatten() - expected_slice).max() < 1e-2, (
             f" expected_slice {expected_slice}, but got {image_slice.flatten()}"

tests/pipelines/kandinsky3/test_kandinsky3_img2img.py

@@ -20,7 +20,7 @@ import unittest
 import numpy as np
 import torch
 from PIL import Image
-from transformers import AutoConfig, AutoTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, T5EncoderModel
 
 from diffusers import (
     AutoPipelineForImage2Image,
@@ -119,8 +119,7 @@ class Kandinsky3Img2ImgPipelineFastTests(PipelineTesterMixin, unittest.TestCase)
         torch.manual_seed(0)
         movq = self.dummy_movq
         torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config).eval()
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         torch.manual_seed(0)
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
@@ -158,7 +157,7 @@ class Kandinsky3Img2ImgPipelineFastTests(PipelineTesterMixin, unittest.TestCase)
     def test_dict_tuple_outputs_equivalent(self):
         expected_slice = None
         if torch_device == "cpu":
-            expected_slice = np.array([0.5261, 0.5688, 0.4093, 0.4865, 0.5326, 0.4480, 0.5064, 0.5113, 0.6222])
+            expected_slice = np.array([0.5762, 0.6112, 0.4150, 0.6018, 0.6167, 0.4626, 0.5426, 0.5641, 0.6536])
         super().test_dict_tuple_outputs_equivalent(expected_slice=expected_slice)
 
     def test_kandinsky3_img2img(self):
@@ -178,7 +177,9 @@ class Kandinsky3Img2ImgPipelineFastTests(PipelineTesterMixin, unittest.TestCase)
 
         assert image.shape == (1, 64, 64, 3)
 
-        expected_slice = np.array([0.5261, 0.5688, 0.4093, 0.4865, 0.5326, 0.4480, 0.5064, 0.5113, 0.6222])
+        expected_slice = np.array(
+            [0.576259, 0.6132097, 0.41703486, 0.603196, 0.62062526, 0.4655338, 0.5434324, 0.5660727, 0.65433365]
+        )
 
         assert np.abs(image_slice.flatten() - expected_slice).max() < 1e-2, (
             f" expected_slice {expected_slice}, but got {image_slice.flatten()}"

tests/pipelines/latte/test_latte.py

@@ -20,7 +20,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import AutoConfig, AutoTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, T5EncoderModel
 
 from diffusers import (
     AutoencoderKL,
@@ -109,8 +109,7 @@ class LattePipelineFastTests(
         vae = AutoencoderKL()
 
         scheduler = DDIMScheduler()
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 

tests/pipelines/ltx/test_ltx.py

@@ -17,7 +17,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import AutoConfig, AutoTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, T5EncoderModel
 
 from diffusers import AutoencoderKLLTXVideo, FlowMatchEulerDiscreteScheduler, LTXPipeline, LTXVideoTransformer3DModel
 
@@ -88,8 +88,7 @@ class LTXPipelineFastTests(PipelineTesterMixin, FirstBlockCacheTesterMixin, unit
 
         torch.manual_seed(0)
         scheduler = FlowMatchEulerDiscreteScheduler()
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         components = {

tests/pipelines/ltx/test_ltx_condition.py

@@ -17,7 +17,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import AutoConfig, AutoTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, T5EncoderModel
 
 from diffusers import (
     AutoencoderKLLTXVideo,
@@ -92,8 +92,7 @@ class LTXConditionPipelineFastTests(PipelineTesterMixin, unittest.TestCase):
 
         torch.manual_seed(0)
         scheduler = FlowMatchEulerDiscreteScheduler()
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         components = {

tests/pipelines/ltx/test_ltx_image2video.py

@@ -17,7 +17,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import AutoConfig, AutoTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, T5EncoderModel
 
 from diffusers import (
     AutoencoderKLLTXVideo,
@@ -91,8 +91,7 @@ class LTXImageToVideoPipelineFastTests(PipelineTesterMixin, unittest.TestCase):
 
         torch.manual_seed(0)
         scheduler = FlowMatchEulerDiscreteScheduler()
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         components = {

tests/pipelines/mochi/test_mochi.py

@@ -18,7 +18,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import AutoConfig, AutoTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, T5EncoderModel
 
 from diffusers import AutoencoderKLMochi, FlowMatchEulerDiscreteScheduler, MochiPipeline, MochiTransformer3DModel
 
@@ -89,8 +89,7 @@ class MochiPipelineFastTests(
 
         torch.manual_seed(0)
         scheduler = FlowMatchEulerDiscreteScheduler()
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         components = {
@@ -208,9 +207,6 @@ class MochiPipelineFastTests(
             return
 
         components = self.get_dummy_components()
-        for key in components:
-            if "text_encoder" in key and hasattr(components[key], "eval"):
-                components[key].eval()
         pipe = self.pipeline_class(**components)
         for component in pipe.components.values():
             if hasattr(component, "set_default_attn_processor"):

tests/pipelines/pag/test_pag_hunyuan_dit.py

@@ -19,7 +19,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import AutoConfig, AutoTokenizer, BertModel, T5EncoderModel
+from transformers import AutoTokenizer, BertModel, T5EncoderModel
 
 from diffusers import (
     AutoencoderKL,
@@ -67,9 +67,7 @@ class HunyuanDiTPAGPipelineFastTests(PipelineTesterMixin, unittest.TestCase):
         scheduler = DDPMScheduler()
         text_encoder = BertModel.from_pretrained("hf-internal-testing/tiny-random-BertModel")
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-BertModel")
-        torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder_2 = T5EncoderModel(config)
+        text_encoder_2 = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer_2 = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         components = {

tests/pipelines/pag/test_pag_pixart_sigma.py

@@ -19,7 +19,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import AutoConfig, AutoTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, T5EncoderModel
 
 import diffusers
 from diffusers import (
@@ -80,8 +80,7 @@ class PixArtSigmaPAGPipelineFastTests(PipelineTesterMixin, unittest.TestCase):
         vae = AutoencoderKL()
 
         scheduler = DDIMScheduler()
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 

tests/pipelines/pag/test_pag_sd3.py

@@ -3,14 +3,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import (
-    AutoConfig,
-    AutoTokenizer,
-    CLIPTextConfig,
-    CLIPTextModelWithProjection,
-    CLIPTokenizer,
-    T5EncoderModel,
-)
+from transformers import AutoTokenizer, CLIPTextConfig, CLIPTextModelWithProjection, CLIPTokenizer, T5EncoderModel
 
 from diffusers import (
     AutoencoderKL,
@@ -80,9 +73,7 @@ class StableDiffusion3PAGPipelineFastTests(unittest.TestCase, PipelineTesterMixi
         torch.manual_seed(0)
         text_encoder_2 = CLIPTextModelWithProjection(clip_text_encoder_config)
 
-        torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder_3 = T5EncoderModel(config)
+        text_encoder_3 = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         tokenizer = CLIPTokenizer.from_pretrained("hf-internal-testing/tiny-random-clip")
         tokenizer_2 = CLIPTokenizer.from_pretrained("hf-internal-testing/tiny-random-clip")

tests/pipelines/pag/test_pag_sd3_img2img.py

@@ -5,14 +5,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import (
-    AutoConfig,
-    AutoTokenizer,
-    CLIPTextConfig,
-    CLIPTextModelWithProjection,
-    CLIPTokenizer,
-    T5EncoderModel,
-)
+from transformers import AutoTokenizer, CLIPTextConfig, CLIPTextModelWithProjection, CLIPTokenizer, T5EncoderModel
 
 from diffusers import (
     AutoencoderKL,
@@ -91,9 +84,7 @@ class StableDiffusion3PAGImg2ImgPipelineFastTests(unittest.TestCase, PipelineTes
         torch.manual_seed(0)
         text_encoder_2 = CLIPTextModelWithProjection(clip_text_encoder_config)
 
-        torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder_3 = T5EncoderModel(config)
+        text_encoder_3 = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         tokenizer = CLIPTokenizer.from_pretrained("hf-internal-testing/tiny-random-clip")
         tokenizer_2 = CLIPTokenizer.from_pretrained("hf-internal-testing/tiny-random-clip")

tests/pipelines/pixart_alpha/test_pixart.py

@@ -19,7 +19,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import AutoConfig, AutoTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, T5EncoderModel
 
 from diffusers import (
     AutoencoderKL,
@@ -77,10 +77,7 @@ class PixArtAlphaPipelineFastTests(PipelineTesterMixin, unittest.TestCase):
         vae = AutoencoderKL()
 
         scheduler = DDIMScheduler()
-
-        torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 

tests/pipelines/pixart_sigma/test_pixart.py

@@ -19,7 +19,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import AutoConfig, AutoTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, T5EncoderModel
 
 from diffusers import (
     AutoencoderKL,
@@ -83,10 +83,7 @@ class PixArtSigmaPipelineFastTests(PipelineTesterMixin, unittest.TestCase):
         vae = AutoencoderKL()
 
         scheduler = DDIMScheduler()
-
-        torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 

tests/pipelines/qwenimage/test_qwenimage.py

@@ -113,7 +113,7 @@ class QwenImagePipelineFastTests(PipelineTesterMixin, unittest.TestCase):
             vision_start_token_id=151652,
             vision_token_id=151654,
         )
-        text_encoder = Qwen2_5_VLForConditionalGeneration(config).eval()
+        text_encoder = Qwen2_5_VLForConditionalGeneration(config)
         tokenizer = Qwen2Tokenizer.from_pretrained("hf-internal-testing/tiny-random-Qwen2VLForConditionalGeneration")
 
         components = {
@@ -160,7 +160,7 @@ class QwenImagePipelineFastTests(PipelineTesterMixin, unittest.TestCase):
         self.assertEqual(generated_image.shape, (3, 32, 32))
 
         # fmt: off
-        expected_slice = torch.tensor([0.5633, 0.6368, 0.6015, 0.5637, 0.5817, 0.5528, 0.5718, 0.6326, 0.4147, 0.3556, 0.5623, 0.4833, 0.4971, 0.5262, 0.4087, 0.5021])
+        expected_slice = torch.tensor([0.56331, 0.63677, 0.6015, 0.56369, 0.58166, 0.55277, 0.57176, 0.63261, 0.41466, 0.35561, 0.56229, 0.48334, 0.49714, 0.52622, 0.40872, 0.50208])
         # fmt: on
 
         generated_slice = generated_image.flatten()

tests/pipelines/qwenimage/test_qwenimage_edit.py

@@ -115,7 +115,7 @@ class QwenImageEditPipelineFastTests(PipelineTesterMixin, unittest.TestCase):
             vision_start_token_id=151652,
             vision_token_id=151654,
         )
-        text_encoder = Qwen2_5_VLForConditionalGeneration(config).eval()
+        text_encoder = Qwen2_5_VLForConditionalGeneration(config)
         tokenizer = Qwen2Tokenizer.from_pretrained(tiny_ckpt_id)
 
         components = {
@@ -163,7 +163,7 @@ class QwenImageEditPipelineFastTests(PipelineTesterMixin, unittest.TestCase):
         self.assertEqual(generated_image.shape, (3, 32, 32))
 
         # fmt: off
-        expected_slice = torch.tensor([0.5637, 0.6341, 0.6001, 0.5620, 0.5794, 0.5498, 0.5757, 0.6389, 0.4174, 0.3597, 0.5649, 0.4894, 0.4969, 0.5255, 0.4083, 0.4986])
+        expected_slice = torch.tensor([[0.5637, 0.6341, 0.6001, 0.5620, 0.5794, 0.5498, 0.5757, 0.6389, 0.4174, 0.3597, 0.5649, 0.4894, 0.4969, 0.5255, 0.4083, 0.4986]])
         # fmt: on
 
         generated_slice = generated_image.flatten()

tests/pipelines/qwenimage/test_qwenimage_edit_plus.py

@@ -164,7 +164,7 @@ class QwenImageEditPlusPipelineFastTests(PipelineTesterMixin, unittest.TestCase)
         self.assertEqual(generated_image.shape, (3, 32, 32))
 
         # fmt: off
-        expected_slice = torch.tensor([0.5640, 0.6339, 0.5997, 0.5607, 0.5799, 0.5496, 0.5760, 0.6393, 0.4172, 0.3595, 0.5655, 0.4896, 0.4971, 0.5255, 0.4088, 0.4987])
+        expected_slice = torch.tensor([[0.5637, 0.6341, 0.6001, 0.5620, 0.5794, 0.5498, 0.5757, 0.6389, 0.4174, 0.3597, 0.5649, 0.4894, 0.4969, 0.5255, 0.4083, 0.4986]])
         # fmt: on
 
         generated_slice = generated_image.flatten()

tests/pipelines/skyreels_v2/test_skyreels_v2.py

@@ -18,11 +18,20 @@ import numpy as np
 import torch
 from transformers import AutoTokenizer, T5EncoderModel
 
-from diffusers import AutoencoderKLWan, SkyReelsV2Pipeline, SkyReelsV2Transformer3DModel, UniPCMultistepScheduler
+from diffusers import (
+    AutoencoderKLWan,
+    SkyReelsV2Pipeline,
+    SkyReelsV2Transformer3DModel,
+    UniPCMultistepScheduler,
+)
 
-from ...testing_utils import enable_full_determinism
+from ...testing_utils import (
+    enable_full_determinism,
+)
 from ..pipeline_params import TEXT_TO_IMAGE_BATCH_PARAMS, TEXT_TO_IMAGE_IMAGE_PARAMS, TEXT_TO_IMAGE_PARAMS
-from ..test_pipelines_common import PipelineTesterMixin
+from ..test_pipelines_common import (
+    PipelineTesterMixin,
+)
 
 
 enable_full_determinism()

tests/pipelines/skyreels_v2/test_skyreels_v2_df.py

@@ -25,9 +25,13 @@ from diffusers import (
     UniPCMultistepScheduler,
 )
 
-from ...testing_utils import enable_full_determinism
+from ...testing_utils import (
+    enable_full_determinism,
+)
 from ..pipeline_params import TEXT_TO_IMAGE_BATCH_PARAMS, TEXT_TO_IMAGE_IMAGE_PARAMS, TEXT_TO_IMAGE_PARAMS
-from ..test_pipelines_common import PipelineTesterMixin
+from ..test_pipelines_common import (
+    PipelineTesterMixin,
+)
 
 
 enable_full_determinism()

tests/pipelines/skyreels_v2/test_skyreels_v2_df_image_to_video.py

@@ -17,7 +17,10 @@ import unittest
 import numpy as np
 import torch
 from PIL import Image
-from transformers import AutoTokenizer, T5EncoderModel
+from transformers import (
+    AutoTokenizer,
+    T5EncoderModel,
+)
 
 from diffusers import (
     AutoencoderKLWan,

tests/pipelines/skyreels_v2/test_skyreels_v2_df_video_to_video.py

@@ -27,9 +27,14 @@ from diffusers import (
     UniPCMultistepScheduler,
 )
 
-from ...testing_utils import enable_full_determinism, torch_device
+from ...testing_utils import (
+    enable_full_determinism,
+    torch_device,
+)
 from ..pipeline_params import TEXT_TO_IMAGE_IMAGE_PARAMS, TEXT_TO_IMAGE_PARAMS
-from ..test_pipelines_common import PipelineTesterMixin
+from ..test_pipelines_common import (
+    PipelineTesterMixin,
+)
 
 
 enable_full_determinism()

tests/pipelines/stable_audio/test_stable_audio.py

@@ -19,7 +19,10 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import AutoConfig, T5EncoderModel, T5Tokenizer
+from transformers import (
+    T5EncoderModel,
+    T5Tokenizer,
+)
 
 from diffusers import (
     AutoencoderOobleck,
@@ -108,8 +111,7 @@ class StableAudioPipelineFastTests(PipelineTesterMixin, unittest.TestCase):
         )
         torch.manual_seed(0)
         t5_repo_id = "hf-internal-testing/tiny-random-T5ForConditionalGeneration"
-        config = AutoConfig.from_pretrained(t5_repo_id)
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained(t5_repo_id)
         tokenizer = T5Tokenizer.from_pretrained(t5_repo_id, truncation=True, model_max_length=25)
 
         torch.manual_seed(0)

tests/pipelines/stable_diffusion_3/test_pipeline_stable_diffusion_3.py

@@ -3,14 +3,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import (
-    AutoConfig,
-    AutoTokenizer,
-    CLIPTextConfig,
-    CLIPTextModelWithProjection,
-    CLIPTokenizer,
-    T5EncoderModel,
-)
+from transformers import AutoTokenizer, CLIPTextConfig, CLIPTextModelWithProjection, CLIPTokenizer, T5EncoderModel
 
 from diffusers import AutoencoderKL, FlowMatchEulerDiscreteScheduler, SD3Transformer2DModel, StableDiffusion3Pipeline
 
@@ -79,9 +72,7 @@ class StableDiffusion3PipelineFastTests(unittest.TestCase, PipelineTesterMixin):
         torch.manual_seed(0)
         text_encoder_2 = CLIPTextModelWithProjection(clip_text_encoder_config)
 
-        torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder_3 = T5EncoderModel(config)
+        text_encoder_3 = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         tokenizer = CLIPTokenizer.from_pretrained("hf-internal-testing/tiny-random-clip")
         tokenizer_2 = CLIPTokenizer.from_pretrained("hf-internal-testing/tiny-random-clip")

tests/pipelines/stable_diffusion_3/test_pipeline_stable_diffusion_3_img2img.py

@@ -4,14 +4,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import (
-    AutoConfig,
-    AutoTokenizer,
-    CLIPTextConfig,
-    CLIPTextModelWithProjection,
-    CLIPTokenizer,
-    T5EncoderModel,
-)
+from transformers import AutoTokenizer, CLIPTextConfig, CLIPTextModelWithProjection, CLIPTokenizer, T5EncoderModel
 
 from diffusers import (
     AutoencoderKL,
@@ -80,9 +73,7 @@ class StableDiffusion3Img2ImgPipelineFastTests(PipelineLatentTesterMixin, unitte
         torch.manual_seed(0)
         text_encoder_2 = CLIPTextModelWithProjection(clip_text_encoder_config)
 
-        torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder_3 = T5EncoderModel(config)
+        text_encoder_3 = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         tokenizer = CLIPTokenizer.from_pretrained("hf-internal-testing/tiny-random-clip")
         tokenizer_2 = CLIPTokenizer.from_pretrained("hf-internal-testing/tiny-random-clip")

tests/pipelines/stable_diffusion_3/test_pipeline_stable_diffusion_3_inpaint.py

@@ -3,14 +3,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import (
-    AutoConfig,
-    AutoTokenizer,
-    CLIPTextConfig,
-    CLIPTextModelWithProjection,
-    CLIPTokenizer,
-    T5EncoderModel,
-)
+from transformers import AutoTokenizer, CLIPTextConfig, CLIPTextModelWithProjection, CLIPTokenizer, T5EncoderModel
 
 from diffusers import (
     AutoencoderKL,
@@ -80,9 +73,7 @@ class StableDiffusion3InpaintPipelineFastTests(PipelineLatentTesterMixin, unitte
         torch.manual_seed(0)
         text_encoder_2 = CLIPTextModelWithProjection(clip_text_encoder_config)
 
-        torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder_3 = T5EncoderModel(config)
+        text_encoder_3 = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         tokenizer = CLIPTokenizer.from_pretrained("hf-internal-testing/tiny-random-clip")
         tokenizer_2 = CLIPTokenizer.from_pretrained("hf-internal-testing/tiny-random-clip")

tests/pipelines/test_pipelines_common.py

@@ -1157,9 +1157,6 @@ class PipelineTesterMixin:
 
     def test_save_load_local(self, expected_max_difference=5e-4):
         components = self.get_dummy_components()
-        for key in components:
-            if "text_encoder" in key and hasattr(components[key], "eval"):
-                components[key].eval()
         pipe = self.pipeline_class(**components)
         for component in pipe.components.values():
             if hasattr(component, "set_default_attn_processor"):
@@ -1298,9 +1295,6 @@ class PipelineTesterMixin:
         additional_params_copy_to_batched_inputs=["num_inference_steps"],
     ):
         components = self.get_dummy_components()
-        for key in components:
-            if "text_encoder" in key and hasattr(components[key], "eval"):
-                components[key].eval()
         pipe = self.pipeline_class(**components)
         for components in pipe.components.values():
             if hasattr(components, "set_default_attn_processor"):
@@ -1351,9 +1345,6 @@ class PipelineTesterMixin:
 
     def test_dict_tuple_outputs_equivalent(self, expected_slice=None, expected_max_difference=1e-4):
         components = self.get_dummy_components()
-        for key in components:
-            if "text_encoder" in key and hasattr(components[key], "eval"):
-                components[key].eval()
         pipe = self.pipeline_class(**components)
         for component in pipe.components.values():
             if hasattr(component, "set_default_attn_processor"):
@@ -1486,9 +1477,6 @@ class PipelineTesterMixin:
         if not self.pipeline_class._optional_components:
             return
         components = self.get_dummy_components()
-        for key in components:
-            if "text_encoder" in key and hasattr(components[key], "eval"):
-                components[key].eval()
         pipe = self.pipeline_class(**components)
         for component in pipe.components.values():
             if hasattr(component, "set_default_attn_processor"):
@@ -1569,9 +1557,6 @@ class PipelineTesterMixin:
             return
 
         components = self.get_dummy_components()
-        for key in components:
-            if "text_encoder" in key and hasattr(components[key], "eval"):
-                components[key].eval()
         pipe = self.pipeline_class(**components)
         for component in pipe.components.values():
             if hasattr(component, "set_default_attn_processor"):
@@ -2080,16 +2065,7 @@ class PipelineTesterMixin:
             for component_name in model_components_pipe:
                 pipe_component = model_components_pipe[component_name]
                 pipe_loaded_component = model_components_pipe_loaded[component_name]
-
-                model_loaded_params = dict(pipe_loaded_component.named_parameters())
-                model_original_params = dict(pipe_component.named_parameters())
-
-                for name, p1 in model_original_params.items():
-                    # Skip tied weights that aren't saved with variants (transformers v5 behavior)
-                    if name not in model_loaded_params:
-                        continue
-
-                    p2 = model_loaded_params[name]
+                for p1, p2 in zip(pipe_component.parameters(), pipe_loaded_component.parameters()):
                     # nan check for luminanext (mps).
                     if not (is_nan(p1) and is_nan(p2)):
                         self.assertTrue(torch.equal(p1, p2))
@@ -2113,9 +2089,6 @@ class PipelineTesterMixin:
             return
 
         components = self.get_dummy_components()
-        for key in components:
-            if "text_encoder" in key and hasattr(components[key], "eval"):
-                components[key].eval()
 
         # We initialize the pipeline with only text encoders and tokenizers,
         # mimicking a real-world scenario.
@@ -2247,9 +2220,6 @@ class PipelineTesterMixin:
         from huggingface_hub import export_folder_as_dduf
 
         components = self.get_dummy_components()
-        for key in components:
-            if "text_encoder" in key and hasattr(components[key], "eval"):
-                components[key].eval()
         pipe = self.pipeline_class(**components)
         pipe = pipe.to(torch_device)
         pipe.set_progress_bar_config(disable=None)
@@ -2387,11 +2357,6 @@ class PipelineTesterMixin:
 
     def test_pipeline_with_accelerator_device_map(self, expected_max_difference=1e-4):
         components = self.get_dummy_components()
-        # Set text encoders to eval mode to match from_pretrained behavior
-        # This ensures deterministic outputs when models are loaded with device_map
-        for key in components:
-            if "text_encoder" in key and hasattr(components[key], "eval"):
-                components[key].eval()
         pipe = self.pipeline_class(**components)
         pipe = pipe.to(torch_device)
         pipe.set_progress_bar_config(disable=None)
@@ -2715,9 +2680,6 @@ class PyramidAttentionBroadcastTesterMixin:
         device = "cpu"  # ensure determinism for the device-dependent torch.Generator
         num_layers = 2
         components = self.get_dummy_components(num_layers=num_layers)
-        for key in components:
-            if "text_encoder" in key and hasattr(components[key], "eval"):
-                components[key].eval()
         pipe = self.pipeline_class(**components)
         pipe = pipe.to(device)
         pipe.set_progress_bar_config(disable=None)

tests/pipelines/visualcloze/test_pipeline_visualcloze_combined.py

@@ -5,7 +5,7 @@ import unittest
 import numpy as np
 import torch
 from PIL import Image
-from transformers import AutoConfig, AutoTokenizer, CLIPTextConfig, CLIPTextModel, CLIPTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, CLIPTextConfig, CLIPTextModel, CLIPTokenizer, T5EncoderModel
 
 import diffusers
 from diffusers import AutoencoderKL, FlowMatchEulerDiscreteScheduler, FluxTransformer2DModel, VisualClozePipeline
@@ -77,8 +77,7 @@ class VisualClozePipelineFastTests(unittest.TestCase, PipelineTesterMixin):
         text_encoder = CLIPTextModel(clip_text_encoder_config)
 
         torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder_2 = T5EncoderModel(config)
+        text_encoder_2 = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         tokenizer = CLIPTokenizer.from_pretrained("hf-internal-testing/tiny-random-clip")
         tokenizer_2 = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
@@ -261,9 +260,6 @@ class VisualClozePipelineFastTests(unittest.TestCase, PipelineTesterMixin):
         if not hasattr(self.pipeline_class, "_optional_components"):
             return
         components = self.get_dummy_components()
-        for key in components:
-            if "text_encoder" in key and hasattr(components[key], "eval"):
-                components[key].eval()
         pipe = self.pipeline_class(**components)
         for component in pipe.components.values():
             if hasattr(component, "set_default_attn_processor"):

tests/pipelines/visualcloze/test_pipeline_visualcloze_generation.py

@@ -5,7 +5,7 @@ import unittest
 import numpy as np
 import torch
 from PIL import Image
-from transformers import AutoConfig, AutoTokenizer, CLIPTextConfig, CLIPTextModel, CLIPTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, CLIPTextConfig, CLIPTextModel, CLIPTokenizer, T5EncoderModel
 
 import diffusers
 from diffusers import (
@@ -79,8 +79,7 @@ class VisualClozeGenerationPipelineFastTests(unittest.TestCase, PipelineTesterMi
         text_encoder = CLIPTextModel(clip_text_encoder_config)
 
         torch.manual_seed(0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder_2 = T5EncoderModel(config)
+        text_encoder_2 = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         tokenizer = CLIPTokenizer.from_pretrained("hf-internal-testing/tiny-random-clip")
         tokenizer_2 = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")

tests/pipelines/wan/test_wan.py

@@ -18,7 +18,7 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import AutoConfig, AutoTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, T5EncoderModel
 
 from diffusers import AutoencoderKLWan, FlowMatchEulerDiscreteScheduler, WanPipeline, WanTransformer3DModel
 
@@ -68,8 +68,7 @@ class WanPipelineFastTests(PipelineTesterMixin, unittest.TestCase):
         torch.manual_seed(0)
         # TODO: impl FlowDPMSolverMultistepScheduler
         scheduler = FlowMatchEulerDiscreteScheduler(shift=7.0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         torch.manual_seed(0)

tests/pipelines/wan/test_wan_22.py

@@ -17,11 +17,14 @@ import unittest
 
 import numpy as np
 import torch
-from transformers import AutoConfig, AutoTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, T5EncoderModel
 
 from diffusers import AutoencoderKLWan, UniPCMultistepScheduler, WanPipeline, WanTransformer3DModel
 
-from ...testing_utils import enable_full_determinism, torch_device
+from ...testing_utils import (
+    enable_full_determinism,
+    torch_device,
+)
 from ..pipeline_params import TEXT_TO_IMAGE_BATCH_PARAMS, TEXT_TO_IMAGE_IMAGE_PARAMS, TEXT_TO_IMAGE_PARAMS
 from ..test_pipelines_common import PipelineTesterMixin
 
@@ -60,8 +63,7 @@ class Wan22PipelineFastTests(PipelineTesterMixin, unittest.TestCase):
 
         torch.manual_seed(0)
         scheduler = UniPCMultistepScheduler(prediction_type="flow_prediction", use_flow_sigmas=True, flow_shift=3.0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         torch.manual_seed(0)
@@ -233,8 +235,7 @@ class Wan225BPipelineFastTests(PipelineTesterMixin, unittest.TestCase):
 
         torch.manual_seed(0)
         scheduler = UniPCMultistepScheduler(prediction_type="flow_prediction", use_flow_sigmas=True, flow_shift=3.0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         torch.manual_seed(0)

tests/pipelines/wan/test_wan_22_image_to_video.py

@@ -18,7 +18,7 @@ import unittest
 import numpy as np
 import torch
 from PIL import Image
-from transformers import AutoConfig, AutoTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, T5EncoderModel
 
 from diffusers import AutoencoderKLWan, UniPCMultistepScheduler, WanImageToVideoPipeline, WanTransformer3DModel
 
@@ -64,8 +64,7 @@ class Wan22ImageToVideoPipelineFastTests(PipelineTesterMixin, unittest.TestCase)
 
         torch.manual_seed(0)
         scheduler = UniPCMultistepScheduler(prediction_type="flow_prediction", use_flow_sigmas=True, flow_shift=3.0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         torch.manual_seed(0)
@@ -249,8 +248,7 @@ class Wan225BImageToVideoPipelineFastTests(PipelineTesterMixin, unittest.TestCas
 
         torch.manual_seed(0)
         scheduler = UniPCMultistepScheduler(prediction_type="flow_prediction", use_flow_sigmas=True, flow_shift=3.0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         torch.manual_seed(0)

tests/pipelines/wan/test_wan_animate.py

@@ -19,7 +19,6 @@ import numpy as np
 import torch
 from PIL import Image
 from transformers import (
-    AutoConfig,
     AutoTokenizer,
     CLIPImageProcessor,
     CLIPVisionConfig,
@@ -79,8 +78,7 @@ class WanAnimatePipelineFastTests(PipelineTesterMixin, unittest.TestCase):
 
         torch.manual_seed(0)
         scheduler = FlowMatchEulerDiscreteScheduler(shift=7.0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         torch.manual_seed(0)

tests/pipelines/wan/test_wan_image_to_video.py

@@ -19,7 +19,6 @@ import numpy as np
 import torch
 from PIL import Image
 from transformers import (
-    AutoConfig,
     AutoTokenizer,
     CLIPImageProcessor,
     CLIPVisionConfig,
@@ -69,8 +68,7 @@ class WanImageToVideoPipelineFastTests(PipelineTesterMixin, unittest.TestCase):
         torch.manual_seed(0)
         # TODO: impl FlowDPMSolverMultistepScheduler
         scheduler = FlowMatchEulerDiscreteScheduler(shift=7.0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         torch.manual_seed(0)
@@ -241,8 +239,7 @@ class WanFLFToVideoPipelineFastTests(PipelineTesterMixin, unittest.TestCase):
         torch.manual_seed(0)
         # TODO: impl FlowDPMSolverMultistepScheduler
         scheduler = FlowMatchEulerDiscreteScheduler(shift=7.0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         torch.manual_seed(0)

tests/pipelines/wan/test_wan_vace.py

@@ -18,7 +18,7 @@ import unittest
 import numpy as np
 import torch
 from PIL import Image
-from transformers import AutoConfig, AutoTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, T5EncoderModel
 
 from diffusers import (
     AutoencoderKLWan,
@@ -67,8 +67,7 @@ class WanVACEPipelineFastTests(PipelineTesterMixin, unittest.TestCase):
 
         torch.manual_seed(0)
         scheduler = FlowMatchEulerDiscreteScheduler(shift=7.0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         torch.manual_seed(0)

tests/pipelines/wan/test_wan_video_to_video.py

@@ -16,7 +16,7 @@ import unittest
 
 import torch
 from PIL import Image
-from transformers import AutoConfig, AutoTokenizer, T5EncoderModel
+from transformers import AutoTokenizer, T5EncoderModel
 
 from diffusers import AutoencoderKLWan, UniPCMultistepScheduler, WanTransformer3DModel, WanVideoToVideoPipeline
 
@@ -62,8 +62,7 @@ class WanVideoToVideoPipelineFastTests(PipelineTesterMixin, unittest.TestCase):
 
         torch.manual_seed(0)
         scheduler = UniPCMultistepScheduler(flow_shift=3.0)
-        config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel.from_pretrained("hf-internal-testing/tiny-random-t5")
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
 
         torch.manual_seed(0)

tests/testing_utils.py

@@ -1357,12 +1357,7 @@ def enable_full_determinism():
     #  variable 'CUDA_LAUNCH_BLOCKING' or 'CUBLAS_WORKSPACE_CONFIG' to be set,
     # depending on the CUDA version, so we set them both here
     os.environ["CUDA_LAUNCH_BLOCKING"] = "1"
-    # Use larger workspace size for PyTorch 2.10+ to avoid CUBLAS_STATUS_NOT_INITIALIZED errors
-    # (catches 2.11 dev versions which report as >= 2.10)
-    if is_torch_version(">=", "2.10"):
-        os.environ["CUBLAS_WORKSPACE_CONFIG"] = ":4096:8"
-    else:
-        os.environ["CUBLAS_WORKSPACE_CONFIG"] = ":16:8"
+    os.environ["CUBLAS_WORKSPACE_CONFIG"] = ":16:8"
     torch.use_deterministic_algorithms(True)
 
     # Enable CUDNN deterministic mode