Skip to content

fix MetaCLIP 2 wrong link & wrong model names in the docstrings #40565

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 5 commits into from
Sep 3, 2025
Merged

fix MetaCLIP 2 wrong link & wrong model names in the docstrings #40565

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
069cc2e
fix MetaCLIP 2 wrong link & wrong model names in the documentation an…
Aug 30, 2025
6abdf78
ruff reformatted
Aug 30, 2025
356e227
update files generated by modular
Aug 31, 2025
4c359cc
update meta_clip2 to metaclip_2 to match the original
Aug 31, 2025
97730a3
_supports_flash_attn = False
Aug 31, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion docs/source/en/model_doc/metaclip_2.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -32,7 +32,7 @@ MetaCLIP 2 is a replication of the original CLIP model trained on 300+ languages
This model was contributed by [nielsr](https://huggingface.co/nielsr).
The original code can be found [here](https://github.com/facebookresearch/MetaCLIP).

You can find all the MetaCLIP 2 checkpoints under the [Meta](https://huggingface.co/facebook?search_models=metaclip-2) organization.
You can find all the MetaCLIP 2 checkpoints under the [Meta](https://huggingface.co/facebook/models?search=metaclip-2) organization.

> [!TIP]
> Click on the MetaCLIP 2 models in the right sidebar for more examples of how to apply MetaCLIP 2 to different image and language tasks.
Expand Down
42 changes: 21 additions & 21 deletions src/transformers/models/metaclip_2/configuration_metaclip_2.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -14,18 +14,18 @@

class MetaClip2TextConfig(PretrainedConfig):
r"""
This is the configuration class to store the configuration of a [`MetaClip2TextModel`]. It is used to instantiate a METACLIP_2
text encoder according to the specified arguments, defining the model architecture. Instantiating a configuration
with the defaults will yield a similar configuration to that of the text encoder of the METACLIP_2
[openai/metaclip_2-vit-base-patch32](https://huggingface.co/openai/metaclip_2-vit-base-patch32) architecture.
This is the configuration class to store the configuration of a [`MetaClip2TextModel`]. It is used to instantiate
a MetaClip2 text encoder according to the specified arguments, defining the model architecture. Instantiating a
configuration with the defaults will yield a similar configuration to that of the MetaClip2
[facebook/metaclip-2-worldwide-huge-quickgelu](https://huggingface.co/facebook/metaclip-2-worldwide-huge-quickgelu) architecture.

Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
documentation from [`PretrainedConfig`] for more information.

Args:
vocab_size (`int`, *optional*, defaults to 49408):
Vocabulary size of the METACLIP_2 text model. Defines the number of different tokens that can be represented by
the `inputs_ids` passed when calling [`MetaClip2Model`].
Vocabulary size of the MetaClip2 text model. Defines the number of different tokens that can be represented by
the `inputs_ids` passed when calling [`MetaClip2TextModel`].
hidden_size (`int`, *optional*, defaults to 512):
Dimensionality of the encoder layers and the pooler layer.
intermediate_size (`int`, *optional*, defaults to 2048):
Expand Down Expand Up @@ -63,10 +63,10 @@ class MetaClip2TextConfig(PretrainedConfig):
```python
>>> from transformers import MetaClip2TextConfig, MetaClip2TextModel

>>> # Initializing a MetaClip2TextConfig with openai/metaclip_2-vit-base-patch32 style configuration
>>> # Initializing a MetaClip2TextConfig with facebook/metaclip-2-worldwide-huge-quickgelu style configuration
>>> configuration = MetaClip2TextConfig()

>>> # Initializing a MetaClip2TextModel (with random weights) from the openai/metaclip_2-vit-base-patch32 style configuration
>>> # Initializing a MetaClip2TextModel (with random weights) from the facebook/metaclip-2-worldwide-huge-quickgelu style configuration
>>> model = MetaClip2TextModel(configuration)

>>> # Accessing the model configuration
Expand Down Expand Up @@ -115,10 +115,10 @@ def __init__(

class MetaClip2VisionConfig(PretrainedConfig):
r"""
This is the configuration class to store the configuration of a [`MetaClip2VisionModel`]. It is used to instantiate a
METACLIP_2 vision encoder according to the specified arguments, defining the model architecture. Instantiating a
configuration with the defaults will yield a similar configuration to that of the vision encoder of the METACLIP_2
[openai/metaclip_2-vit-base-patch32](https://huggingface.co/openai/metaclip_2-vit-base-patch32) architecture.
This is the configuration class to store the configuration of a [`MetaClip2VisionModel`]. It is used to instantiate a MetaClip2
vision encoder according to the specified arguments, defining the model architecture. Instantiating a configuration
with the defaults will yield a similar configuration to that of the vision encoder of the MetaClip2
[facebook/metaclip-2-worldwide-huge-quickgelu](https://huggingface.co/facebook/metaclip-2-worldwide-huge-quickgelu) architecture.

Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
documentation from [`PretrainedConfig`] for more information.
Expand Down Expand Up @@ -158,10 +158,10 @@ class MetaClip2VisionConfig(PretrainedConfig):
```python
>>> from transformers import MetaClip2VisionConfig, MetaClip2VisionModel

>>> # Initializing a MetaClip2VisionConfig with openai/metaclip_2-vit-base-patch32 style configuration
>>> # Initializing a MetaClip2VisionConfig with facebook/metaclip-2-worldwide-huge-quickgelu style configuration
>>> configuration = MetaClip2VisionConfig()

>>> # Initializing a MetaClip2VisionModel (with random weights) from the openai/metaclip_2-vit-base-patch32 style configuration
>>> # Initializing a MetaClip2VisionModel (with random weights) from the facebook/metaclip-2-worldwide-huge-quickgelu style configuration
>>> model = MetaClip2VisionModel(configuration)

>>> # Accessing the model configuration
Expand Down Expand Up @@ -207,10 +207,10 @@ def __init__(

class MetaClip2Config(PretrainedConfig):
r"""
[`MetaClip2Config`] is the configuration class to store the configuration of a [`MetaClip2Model`]. It is used to instantiate
a METACLIP_2 model according to the specified arguments, defining the text model and vision model configs. Instantiating
a configuration with the defaults will yield a similar configuration to that of the METACLIP_2
[openai/metaclip_2-vit-base-patch32](https://huggingface.co/openai/metaclip_2-vit-base-patch32) architecture.
[`MetaClip2Config`] is the configuration class to store the configuration of a [`MetaClip2Model`]. It is used to
instantiate a MetaClip2 model according to the specified arguments, defining the text model and vision model configs.
Instantiating a configuration with the defaults will yield a similar configuration to that of the MetaClip2
[facebook/metaclip-2-worldwide-huge-quickgelu](https://huggingface.co/facebook/metaclip-2-worldwide-huge-quickgelu) architecture.

Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
documentation from [`PretrainedConfig`] for more information.
Expand All @@ -223,7 +223,7 @@ class MetaClip2Config(PretrainedConfig):
projection_dim (`int`, *optional*, defaults to 512):
Dimensionality of text and vision projection layers.
logit_scale_init_value (`float`, *optional*, defaults to 2.6592):
The initial value of the *logit_scale* parameter. Default is used as per the original METACLIP_2 implementation.
The initial value of the *logit_scale* parameter. Default is used as per the original MetaClip2 implementation.
kwargs (*optional*):
Dictionary of keyword arguments.

Expand All @@ -232,10 +232,10 @@ class MetaClip2Config(PretrainedConfig):
```python
>>> from transformers import MetaClip2Config, MetaClip2Model

>>> # Initializing a MetaClip2Config with openai/metaclip_2-vit-base-patch32 style configuration
>>> # Initializing a MetaClip2Config with facebook/metaclip-2-worldwide-huge-quickgelu style configuration
>>> configuration = MetaClip2Config()

>>> # Initializing a MetaClip2Model (with random weights) from the openai/metaclip_2-vit-base-patch32 style configuration
>>> # Initializing a MetaClip2Model (with random weights) from the facebook/metaclip-2-worldwide-huge-quickgelu style configuration
>>> model = MetaClip2Model(configuration)

>>> # Accessing the model configuration
Expand Down
Loading