<!-- # hard line break macro for HTML -->

# fiftyone.utils.winnow

[Winnow](https://huggingface.co/EldanRing/Winnow-12B) decision model wrapper
for the FiftyOne Model Zoo.

Copyright 2017-2026, Voxel51, Inc.
<br/>
[voxel51.com](https://voxel51.com/)
<br/>
<br/>

**Classes:**

| [`WinnowGetItem`](#fiftyone.utils.winnow.WinnowGetItem)(processor, prefix[, field_mapping])      | A [`fiftyone.utils.torch.GetItem`](fiftyone.utils.torch.md#fiftyone.utils.torch.GetItem) that loads an image and encodes it into the prompt prefix that every question continues.   |
|--------------------------------------------------------------------------------------------------|-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| [`WinnowOutputProcessor`](#fiftyone.utils.winnow.WinnowOutputProcessor)([classes, store_logits]) | Output processor for Winnow decision models.                                                                                                                                        |
| [`WinnowModelConfig`](#fiftyone.utils.winnow.WinnowModelConfig)(d)                               | Configuration for running a [`WinnowModel`](#fiftyone.utils.winnow.WinnowModel).                                                                                                    |
| [`WinnowModel`](#fiftyone.utils.winnow.WinnowModel)(config)                                      | Wrapper for running inference with Winnow decision models.                                                                                                                          |

### *class* fiftyone.utils.winnow.WinnowGetItem(processor, prefix, field_mapping=None, \*\*kwargs)

Bases: [`GetItem`](fiftyone.utils.torch.md#fiftyone.utils.torch.GetItem)

A [`fiftyone.utils.torch.GetItem`](fiftyone.utils.torch.md#fiftyone.utils.torch.GetItem) that loads an image and encodes
it into the prompt prefix that every question continues.

Data loader workers call it, so image decoding, patching and
tokenization run in parallel with inference.

* **Parameters:**
  * **processor** – the release’s processor
  * **prefix** – the prompt text before the question, holding the image
    placeholder and the state
  * **field_mapping** (*None*) – the user-supplied dict mapping keys in
    [`required_keys`](#fiftyone.utils.winnow.WinnowGetItem.required_keys) to field names of their dataset that
    contain the required values

**Attributes:**

| [`required_keys`](#fiftyone.utils.winnow.WinnowGetItem.required_keys)   | The sample's `filepath`, which locates its image.                                                                                                                                   |
|-------------------------------------------------------------------------|-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| [`field_mapping`](#fiftyone.utils.winnow.WinnowGetItem.field_mapping)   | A user-supplied dictionary mappings keys in [`required_keys`](#fiftyone.utils.winnow.WinnowGetItem.required_keys) to field names of their dataset that contain the required values. |

**Methods:**

| [`encode`](#fiftyone.utils.winnow.WinnowGetItem.encode)(img)   | Encodes an image into the prompt prefix, at the size the release's llama.cpp server gives it.   |
|----------------------------------------------------------------|-------------------------------------------------------------------------------------------------|

#### *property* required_keys

The sample’s `filepath`, which locates its image.

#### encode(img)

Encodes an image into the prompt prefix, at the size the release’s
llama.cpp server gives it.

* **Parameters:**
  **img** – an image path, PIL image, numpy array or Torch tensor
* **Returns:**
  a dict with the prefix’s `input_ids` and `mm_token_type_ids`
  and the image’s `pixel_values` and `image_position_ids`

#### *property* field_mapping

A user-supplied dictionary mappings keys in [`required_keys`](#fiftyone.utils.winnow.WinnowGetItem.required_keys)
to field names of their dataset that contain the required values.

### *class* fiftyone.utils.winnow.WinnowOutputProcessor(classes=None, store_logits=False, \*\*kwargs)

Bases: [`OutputProcessor`](fiftyone.utils.torch.md#fiftyone.utils.torch.OutputProcessor)

Output processor for Winnow decision models.

A `choice` or `noul` answer becomes a
[`fiftyone.core.labels.Classification`](fiftyone.core.labels.md#fiftyone.core.labels.Classification) whose label is the most
probable option and whose confidence is that option’s probability. A
`score` answer becomes a [`fiftyone.core.labels.Regression`](fiftyone.core.labels.md#fiftyone.core.labels.Regression) whose
value is the expected level and whose confidence is the probability of
the most probable level. A schema of one question yields that label; a
schema of several yields a dict keyed by question ID.

* **Parameters:**
  * **classes** (*None*) – unused; every answer carries its own options
  * **store_logits** (*False*) – whether to store each classification’s
    per-option logits, in the order the options were given

### *class* fiftyone.utils.winnow.WinnowModelConfig(d)

Bases: [`TorchImageModelConfig`](fiftyone.utils.torch.md#fiftyone.utils.torch.TorchImageModelConfig), [`HasZooModel`](fiftyone.zoo.models.md#fiftyone.zoo.models.HasZooModel)

Configuration for running a [`WinnowModel`](#fiftyone.utils.winnow.WinnowModel).

Exactly one of `questions` and `classes` must be provided.

* **Parameters:**
  * **name_or_path** ( *"EldanRing/Winnow-12B"*) – the HuggingFace repository of
    a Winnow release
  * **revision** (*None*) – the commit hash, branch or tag of the release to
    load. The wrapper reads the release’s files by name, so the
    default release loads its pinned revision when this is omitted
  * **questions** (*None*) – 

    the questions to answer about each image, a dict
    mapping a question ID, which must be a valid field name, to a
    dict with:
    - `type`: `"noul"` (yes or no), `"choice"` (one of 2 to
      64 named options) or `"score"` (one of 2 to 64 ordered
      levels)
    - `instructions`: what to decide; the question ID is used
      when omitted
    - `criteria`: for `choice`, a dict mapping each option to
      its description; for `score`, a list of level
      descriptions, lowest first; for `noul`, optional
      descriptions of `"true"` and `"false"`
  * **classes** (*None*) – the options of a single `choice` question, for
    zero-shot classification
  * **prompt** (*None*) – the instructions of the question built from
    `classes`; if None, asks which option best describes the image
  * **state** ( *"The attached image."*) – the text or JSON value describing the
    situation, given with every image

**Methods:**

| [`get_questions`](#fiftyone.utils.winnow.WinnowModelConfig.get_questions)()                                           | Returns the question schema the model answers for each image.                                                          |
|-----------------------------------------------------------------------------------------------------------------------|------------------------------------------------------------------------------------------------------------------------|
| [`attributes`](#fiftyone.utils.winnow.WinnowModelConfig.attributes)()                                                 | Returns a list of class attributes to be serialized.                                                                   |
| [`builder`](#fiftyone.utils.winnow.WinnowModelConfig.builder)()                                                       | Returns a ConfigBuilder instance for this class.                                                                       |
| [`copy`](#fiftyone.utils.winnow.WinnowModelConfig.copy)()                                                             | Returns a deep copy of the object.                                                                                     |
| [`custom_attributes`](#fiftyone.utils.winnow.WinnowModelConfig.custom_attributes)([dynamic, private])                 | Returns a customizable list of class attributes.                                                                       |
| [`default`](#fiftyone.utils.winnow.WinnowModelConfig.default)()                                                       | Returns the default config instance.                                                                                   |
| [`download_model_if_necessary`](#fiftyone.utils.winnow.WinnowModelConfig.download_model_if_necessary)()               | Downloads the published model specified by the config, if necessary.                                                   |
| [`from_dict`](#fiftyone.utils.winnow.WinnowModelConfig.from_dict)(d)                                                  | Constructs a Config object from a JSON dictionary.                                                                     |
| [`from_json`](#fiftyone.utils.winnow.WinnowModelConfig.from_json)(path, \*args, \*\*kwargs)                           | Constructs a Serializable object from a JSON file.                                                                     |
| [`from_kwargs`](#fiftyone.utils.winnow.WinnowModelConfig.from_kwargs)(\*\*kwargs)                                     | Constructs a Config object from keyword arguments.                                                                     |
| [`from_str`](#fiftyone.utils.winnow.WinnowModelConfig.from_str)(s, \*args, \*\*kwargs)                                | Constructs a Serializable object from a JSON string.                                                                   |
| [`get_class_name`](#fiftyone.utils.winnow.WinnowModelConfig.get_class_name)()                                         | Returns the fully-qualified class name string of this object.                                                          |
| [`init`](#fiftyone.utils.winnow.WinnowModelConfig.init)(d)                                                            | Initializes the published model config.                                                                                |
| [`load_default`](#fiftyone.utils.winnow.WinnowModelConfig.load_default)()                                             | Loads the default config instance from file.                                                                           |
| [`parse_array`](#fiftyone.utils.winnow.WinnowModelConfig.parse_array)(d, key[, default])                              | Parses a raw array attribute.                                                                                          |
| [`parse_bool`](#fiftyone.utils.winnow.WinnowModelConfig.parse_bool)(d, key[, default])                                | Parses a boolean value.                                                                                                |
| [`parse_categorical`](#fiftyone.utils.winnow.WinnowModelConfig.parse_categorical)(d, key, choices[, default])         | Parses a categorical JSON field, which must take a value from among the given choices.                                 |
| [`parse_dict`](#fiftyone.utils.winnow.WinnowModelConfig.parse_dict)(d, key[, default])                                | Parses a dictionary attribute.                                                                                         |
| [`parse_int`](#fiftyone.utils.winnow.WinnowModelConfig.parse_int)(d, key[, default])                                  | Parses an integer attribute.                                                                                           |
| [`parse_mutually_exclusive_fields`](#fiftyone.utils.winnow.WinnowModelConfig.parse_mutually_exclusive_fields)(fields) | Parses a mutually exclusive dictionary of pre-parsed fields, which must contain exactly one field with a truthy value. |
| [`parse_number`](#fiftyone.utils.winnow.WinnowModelConfig.parse_number)(d, key[, default])                            | Parses a number attribute.                                                                                             |
| [`parse_object`](#fiftyone.utils.winnow.WinnowModelConfig.parse_object)(d, key, cls[, default])                       | Parses an object attribute.                                                                                            |
| [`parse_object_array`](#fiftyone.utils.winnow.WinnowModelConfig.parse_object_array)(d, key, cls[, default])           | Parses an array of objects.                                                                                            |
| [`parse_object_dict`](#fiftyone.utils.winnow.WinnowModelConfig.parse_object_dict)(d, key, cls[, default])             | Parses a dictionary whose values are objects.                                                                          |
| [`parse_path`](#fiftyone.utils.winnow.WinnowModelConfig.parse_path)(d, key[, default])                                | Parses a path attribute.                                                                                               |
| [`parse_raw`](#fiftyone.utils.winnow.WinnowModelConfig.parse_raw)(d, key[, default])                                  | Parses a raw (arbitrary) JSON field.                                                                                   |
| [`parse_string`](#fiftyone.utils.winnow.WinnowModelConfig.parse_string)(d, key[, default])                            | Parses a string attribute.                                                                                             |
| [`serialize`](#fiftyone.utils.winnow.WinnowModelConfig.serialize)([reflective])                                       | Serializes the object into a dictionary.                                                                               |
| [`to_str`](#fiftyone.utils.winnow.WinnowModelConfig.to_str)([pretty_print])                                           | Returns a string representation of this object.                                                                        |
| [`validate_all_or_nothing_fields`](#fiftyone.utils.winnow.WinnowModelConfig.validate_all_or_nothing_fields)(fields)   | Validates a dictionary of pre-parsed fields checking that either all or none of the fields have a truthy value.        |
| [`write_json`](#fiftyone.utils.winnow.WinnowModelConfig.write_json)(path[, pretty_print])                             | Serializes the object and writes it to disk.                                                                           |

#### get_questions()

Returns the question schema the model answers for each image.

* **Returns:**
  a dict mapping question ID to question

#### attributes()

Returns a list of class attributes to be serialized.

This method is called internally by `serialize()` to determine the
class attributes to serialize.

Subclasses can override this method, but, by default, all attributes in
vars(self) are returned, minus private attributes, i.e., those starting
with “_”. The order of the attributes in this list is preserved when
serializing objects, so a common pattern is for subclasses to override
this method if they want their JSON files to be organized in a
particular way.

* **Returns:**
  a list of class attributes to be serialized

#### *classmethod* builder()

Returns a ConfigBuilder instance for this class.

#### copy()

Returns a deep copy of the object.

* **Returns:**
  a Serializable instance

#### custom_attributes(dynamic=False, private=False)

Returns a customizable list of class attributes.

By default, all attributes in vars(self) are returned, minus private
attributes (those starting with “_”).

* **Parameters:**
  * **dynamic** – whether to include dynamic properties, e.g., those defined
    by getter/setter methods or the `@property` decorator. By
    default, this is False
  * **private** – whether to include private properties, i.e., those
    starting with “_”. By default, this is False
* **Returns:**
  a list of class attributes

#### *classmethod* default()

Returns the default config instance.

By default, this method instantiates the class from an empty
dictionary, which will only succeed if all attributes are optional.
Otherwise, subclasses should override this method to provide the
desired default configuration.

#### download_model_if_necessary()

Downloads the published model specified by the config, if necessary.

After this method is called, the `model_path` attribute will always
contain the path to the model on disk.

#### *classmethod* from_dict(d)

Constructs a Config object from a JSON dictionary.

Config subclass constructors accept JSON dictionaries, so this method
simply passes the dictionary to cls().

* **Parameters:**
  **d** – a dict of fields expected by cls
* **Returns:**
  an instance of cls

#### *classmethod* from_json(path, \*args, \*\*kwargs)

Constructs a Serializable object from a JSON file.

Subclasses may override this method, but, by default, this method
simply reads the JSON and calls from_dict(), which subclasses must
implement.

* **Parameters:**
  * **path** – the path to the JSON file on disk
  * **\*args** – optional positional arguments for `self.from_dict()`
  * **\*\*kwargs** – optional keyword arguments for `self.from_dict()`
* **Returns:**
  an instance of the Serializable class

#### *classmethod* from_kwargs(\*\*kwargs)

Constructs a Config object from keyword arguments.

* **Parameters:**
  **\*\*kwargs** – keyword arguments that define the fields expected by cls
* **Returns:**
  an instance of cls

#### *classmethod* from_str(s, \*args, \*\*kwargs)

Constructs a Serializable object from a JSON string.

Subclasses may override this method, but, by default, this method
simply parses the string and calls from_dict(), which subclasses must
implement.

* **Parameters:**
  * **s** – a JSON string representation of a Serializable object
  * **\*args** – optional positional arguments for `self.from_dict()`
  * **\*\*kwargs** – optional keyword arguments for `self.from_dict()`
* **Returns:**
  an instance of the Serializable class

#### *classmethod* get_class_name()

Returns the fully-qualified class name string of this object.

#### init(d)

Initializes the published model config.

This method should be called by `ModelConfig.__init__()`, and it
performs the following tasks:

- Parses the `model_name` and `model_path` parameters
- Populates any default parameters in the provided ModelConfig dict

* **Parameters:**
  **d** – a ModelConfig dict
* **Returns:**
  a ModelConfig dict with any default parameters populated

#### *classmethod* load_default()

Loads the default config instance from file.

Subclasses must implement this method if they intend to support
default instances.

#### *static* parse_array(d, key, default=<eta.core.config.NoDefault object>)

Parses a raw array attribute.

* **Parameters:**
  * **d** – a JSON dictionary
  * **key** – the key to parse
  * **default** – a default list to return if key is not present
* **Returns:**
  a list of raw (untouched) values
* **Raises:**
  [**ConfigError**](fiftyone.zoo.md#fiftyone.zoo.ConfigError) – if the field value was the wrong type or no default
      value was provided and the key was not found in the dictionary

#### *static* parse_bool(d, key, default=<eta.core.config.NoDefault object>)

Parses a boolean value.

* **Parameters:**
  * **d** – a JSON dictionary
  * **key** – the key to parse
  * **default** – a default bool to return if key is not present
* **Returns:**
  True/False
* **Raises:**
  [**ConfigError**](fiftyone.zoo.md#fiftyone.zoo.ConfigError) – if the field value was the wrong type or no default
      value was provided and the key was not found in the dictionary

#### *static* parse_categorical(d, key, choices, default=<eta.core.config.NoDefault object>)

Parses a categorical JSON field, which must take a value from among
the given choices.

* **Parameters:**
  * **d** – a JSON dictionary
  * **key** – the key to parse
  * **choices** – either an iterable of possible values or an enum-like
    class whose attributes define the possible values
  * **default** – a default value to return if key is not present
* **Returns:**
  the raw (untouched) value of the given field, which is equal to a
  value from `choices`
* **Raises:**
  [**ConfigError**](fiftyone.zoo.md#fiftyone.zoo.ConfigError) – if the key was present in the dictionary but its value
      was not an allowed choice, or if no default value was provided
      and the key was not found in the dictionary

#### *static* parse_dict(d, key, default=<eta.core.config.NoDefault object>)

Parses a dictionary attribute.

* **Parameters:**
  * **d** – a JSON dictionary
  * **key** – the key to parse
  * **default** – a default dict to return if key is not present
* **Returns:**
  a dictionary
* **Raises:**
  [**ConfigError**](fiftyone.zoo.md#fiftyone.zoo.ConfigError) – if the field value was the wrong type or no default
      value was provided and the key was not found in the dictionary

#### *static* parse_int(d, key, default=<eta.core.config.NoDefault object>)

Parses an integer attribute.

* **Parameters:**
  * **d** – a JSON dictionary
  * **key** – the key to parse
  * **default** – a default integer value to return if key is not present
* **Returns:**
  an int
* **Raises:**
  [**ConfigError**](fiftyone.zoo.md#fiftyone.zoo.ConfigError) – if the field value was the wrong type or no default
      value was provided and the key was not found in the dictionary

#### *static* parse_mutually_exclusive_fields(fields)

Parses a mutually exclusive dictionary of pre-parsed fields, which
must contain exactly one field with a truthy value.

* **Parameters:**
  **fields** – a dictionary of pre-parsed fields
* **Returns:**
  the (field, value) that was set
* **Raises:**
  [**ConfigError**](fiftyone.zoo.md#fiftyone.zoo.ConfigError) – if zero or more than one truthy value was found

#### *static* parse_number(d, key, default=<eta.core.config.NoDefault object>)

Parses a number attribute.

* **Parameters:**
  * **d** – a JSON dictionary
  * **key** – the key to parse
  * **default** – a default numeric value to return if key is not present
* **Returns:**
  a number (e.g. int, float)
* **Raises:**
  [**ConfigError**](fiftyone.zoo.md#fiftyone.zoo.ConfigError) – if the field value was the wrong type or no default
      value was provided and the key was not found in the dictionary

#### *static* parse_object(d, key, cls, default=<eta.core.config.NoDefault object>)

Parses an object attribute.

The value of d[key] can be either an instance of cls or a serialized
dict from an instance of cls.

* **Parameters:**
  * **d** – a JSON dictionary
  * **key** – the key to parse
  * **cls** – the class of d[key]
  * **default** – a default cls instance to return if key is not present
* **Returns:**
  an instance of cls
* **Raises:**
  [**ConfigError**](fiftyone.zoo.md#fiftyone.zoo.ConfigError) – if the field value was the wrong type or no default
      value was provided and the key was not found in the dictionary

#### *static* parse_object_array(d, key, cls, default=<eta.core.config.NoDefault object>)

Parses an array of objects.

The values in d[key] can be either instances of cls or serialized
dicts from instances of cls.

* **Parameters:**
  * **d** – a JSON dictionary
  * **key** – the key to parse
  * **cls** – the class of the elements of list d[key]
  * **default** – the default list to return if key is not present
* **Returns:**
  a list of cls instances
* **Raises:**
  [**ConfigError**](fiftyone.zoo.md#fiftyone.zoo.ConfigError) – if the field value was the wrong type or no default
      value was provided and the key was not found in the dictionary

#### *static* parse_object_dict(d, key, cls, default=<eta.core.config.NoDefault object>)

Parses a dictionary whose values are objects.

The values in d[key] can be either instances of cls or serialized
dicts from instances of cls.

* **Parameters:**
  * **d** – a JSON dictionary
  * **key** – the key to parse
  * **cls** – the class of the values of dictionary d[key]
  * **default** – the default dict of cls instances to return if key is not
    present
* **Returns:**
  a dictionary whose values are cls instances
* **Raises:**
  [**ConfigError**](fiftyone.zoo.md#fiftyone.zoo.ConfigError) – if the field value was the wrong type or no default
      value was provided and the key was not found in the dictionary

#### *static* parse_path(d, key, default=<eta.core.config.NoDefault object>)

Parses a path attribute.

The path is converted to an absolute path if necessary via
`os.path.abspath(os.path.expanduser(value))`.

* **Parameters:**
  * **d** – a JSON dictionary
  * **key** – the key to parse
  * **default** – a default string to return if key is not present
* **Returns:**
  a path string
* **Raises:**
  [**ConfigError**](fiftyone.zoo.md#fiftyone.zoo.ConfigError) – if the field value was the wrong type or no default
      value was provided and the key was not found in the dictionary

#### *static* parse_raw(d, key, default=<eta.core.config.NoDefault object>)

Parses a raw (arbitrary) JSON field.

* **Parameters:**
  * **d** – a JSON dictionary
  * **key** – the key to parse
  * **default** – a default value to return if key is not present
* **Returns:**
  the raw (untouched) value of the given field
* **Raises:**
  [**ConfigError**](fiftyone.zoo.md#fiftyone.zoo.ConfigError) – if no default value was provided and the key was not
      found in the dictionary

#### *static* parse_string(d, key, default=<eta.core.config.NoDefault object>)

Parses a string attribute.

* **Parameters:**
  * **d** – a JSON dictionary
  * **key** – the key to parse
  * **default** – a default string to return if key is not present
* **Returns:**
  a string
* **Raises:**
  [**ConfigError**](fiftyone.zoo.md#fiftyone.zoo.ConfigError) – if the field value was the wrong type or no default
      value was provided and the key was not found in the dictionary

#### serialize(reflective=False)

Serializes the object into a dictionary.

Serialization is applied recursively to all attributes in the object,
including element-wise serialization of lists and dictionary values.

* **Parameters:**
  **reflective** – whether to include reflective attributes when
  serializing the object. By default, this is False
* **Returns:**
  a JSON dictionary representation of the object

#### to_str(pretty_print=True, \*\*kwargs)

Returns a string representation of this object.

* **Parameters:**
  * **pretty_print** – whether to render the JSON in human readable format
    with newlines and indentations. By default, this is True
  * **\*\*kwargs** – optional keyword arguments for `self.serialize()`
* **Returns:**
  a string representation of the object

#### *static* validate_all_or_nothing_fields(fields)

Validates a dictionary of pre-parsed fields checking that either
all or none of the fields have a truthy value.

* **Parameters:**
  **fields** – a dictionary of pre-parsed fields
* **Raises:**
  [**ConfigError**](fiftyone.zoo.md#fiftyone.zoo.ConfigError) – if some values are truth and some are not

#### write_json(path, pretty_print=False, \*\*kwargs)

Serializes the object and writes it to disk.

* **Parameters:**
  * **path** – the output path
  * **pretty_print** – whether to render the JSON in human readable format
    with newlines and indentations. By default, this is False
  * **\*\*kwargs** – optional keyword arguments for `self.serialize()`

### *class* fiftyone.utils.winnow.WinnowModel(config)

Bases: [`TorchImageModel`](fiftyone.utils.torch.md#fiftyone.utils.torch.TorchImageModel)

Wrapper for running inference with Winnow decision models.

A Winnow model answers typed questions about each image in one forward
pass, reading each answer from the logits of its options’ letter labels
rather than generating text. The release publishes its weights in
llama.cpp’s GGUF format; this wrapper reads them into transformers’
implementation of Gemma 4, sizes each image as the release’s llama.cpp
server does and builds the prompt that its decision server builds.

Zero-shot classification example:

```default
import fiftyone as fo
import fiftyone.zoo as foz

dataset = foz.load_zoo_dataset("quickstart", max_samples=25)

model = foz.load_zoo_model(
    "winnow-12b-torch",
    classes=["indoor", "outdoor"],
    prompt="Was this photo taken indoors or outdoors?",
)

dataset.apply_model(model, label_field="setting", batch_size=8)

session = fo.launch_app(dataset)
```

Several questions at once, one field per question:

```default
model = foz.load_zoo_model(
    "winnow-12b-torch",
    questions={
        "has_person": {
            "type": "noul",
            "instructions": "Is there a person in the image?",
        },
        "quality": {
            "type": "score",
            "instructions": "How sharp is the image?",
            "criteria": ["blurry", "acceptable", "sharp"],
        },
    },
)

# Populates `winnow_has_person` and `winnow_quality`
dataset.apply_model(model, label_field="winnow", batch_size=8)
```

* **Parameters:**
  **config** – a [`WinnowModelConfig`](#fiftyone.utils.winnow.WinnowModelConfig)

**Attributes:**

| [`has_logits`](#fiftyone.utils.winnow.WinnowModel.has_logits)                     | True; classifications can store their per-option logits.                                                                     |
|-----------------------------------------------------------------------------------|------------------------------------------------------------------------------------------------------------------------------|
| [`questions`](#fiftyone.utils.winnow.WinnowModel.questions)                       | A copy of the question schema answered for each image.                                                                       |
| [`ragged_batches`](#fiftyone.utils.winnow.WinnowModel.ragged_batches)             | False; [`collate_fn`](#fiftyone.utils.winnow.WinnowModel.collate_fn) pads rows of different lengths.                         |
| [`has_collate_fn`](#fiftyone.utils.winnow.WinnowModel.has_collate_fn)             | True; the model pads its own batches.                                                                                        |
| [`collate_fn`](#fiftyone.utils.winnow.WinnowModel.collate_fn)                     | The function that continues each image's prompt prefix with every question and pads the rows.                                |
| [`can_embed_prompts`](#fiftyone.utils.winnow.WinnowModel.can_embed_prompts)       | Whether this model can generate prompt embeddings.                                                                           |
| [`classes`](#fiftyone.utils.winnow.WinnowModel.classes)                           | The list of class labels for the model, if known.                                                                            |
| [`device`](#fiftyone.utils.winnow.WinnowModel.device)                             | The `torch:torch.torch.device` that the model is using.                                                                      |
| [`has_embeddings`](#fiftyone.utils.winnow.WinnowModel.has_embeddings)             | Whether this model has embeddings.                                                                                           |
| [`mask_targets`](#fiftyone.utils.winnow.WinnowModel.mask_targets)                 | The mask targets for the model, if any.                                                                                      |
| [`media_type`](#fiftyone.utils.winnow.WinnowModel.media_type)                     | The media type processed by the model.                                                                                       |
| [`num_classes`](#fiftyone.utils.winnow.WinnowModel.num_classes)                   | The number of classes for the model, if known.                                                                               |
| [`preprocess`](#fiftyone.utils.winnow.WinnowModel.preprocess)                     | Whether to apply preprocessing transforms for inference, if any.                                                             |
| [`required_keys`](#fiftyone.utils.winnow.WinnowModel.required_keys)               | The required keys that must be provided as parameters to methods like `apply_model()` and `compute_embeddings()` at runtime. |
| [`skeleton`](#fiftyone.utils.winnow.WinnowModel.skeleton)                         | The keypoint skeleton for the model, if any.                                                                                 |
| [`store_logits`](#fiftyone.utils.winnow.WinnowModel.store_logits)                 | Whether the model should store logits in its predictions.                                                                    |
| [`transforms`](#fiftyone.utils.winnow.WinnowModel.transforms)                     | A `torchvision.transforms` function that will be applied to each input before prediction, if any.                            |
| [`using_gpu`](#fiftyone.utils.winnow.WinnowModel.using_gpu)                       | Whether the model is using GPU.                                                                                              |
| [`using_half_precision`](#fiftyone.utils.winnow.WinnowModel.using_half_precision) | Whether the model is using half precision.                                                                                   |

**Methods:**

| [`build_get_item`](#fiftyone.utils.winnow.WinnowModel.build_get_item)([field_mapping])   | Builds the [`WinnowGetItem`](#fiftyone.utils.winnow.WinnowGetItem) that encodes each image into this model's prompt prefix.   |
|------------------------------------------------------------------------------------------|-------------------------------------------------------------------------------------------------------------------------------|
| [`embed`](#fiftyone.utils.winnow.WinnowModel.embed)(arg)                                 | Generates an embedding for the given data.                                                                                    |
| [`embed_all`](#fiftyone.utils.winnow.WinnowModel.embed_all)(args)                        | Generates embeddings for the given iterable of data.                                                                          |
| [`from_config`](#fiftyone.utils.winnow.WinnowModel.from_config)(config)                  | Instantiates a Configurable class from a <cls>Config instance.                                                                |
| [`from_dict`](#fiftyone.utils.winnow.WinnowModel.from_dict)(d)                           | Instantiates a Configurable class from a <cls>Config dict.                                                                    |
| [`from_json`](#fiftyone.utils.winnow.WinnowModel.from_json)(json_path)                   | Instantiates a Configurable class from a <cls>Config JSON file.                                                               |
| [`from_kwargs`](#fiftyone.utils.winnow.WinnowModel.from_kwargs)(\*\*kwargs)              | Instantiates a Configurable class from keyword arguments defining the attributes of a <cls>Config.                            |
| [`get_embeddings`](#fiftyone.utils.winnow.WinnowModel.get_embeddings)()                  | Returns the embeddings generated by the last forward pass of the model.                                                       |
| [`parse`](#fiftyone.utils.winnow.WinnowModel.parse)(class_name[, module_name])           | Parses a Configurable subclass name string.                                                                                   |
| [`predict`](#fiftyone.utils.winnow.WinnowModel.predict)(img)                             | Performs prediction on the given image.                                                                                       |
| [`predict_all`](#fiftyone.utils.winnow.WinnowModel.predict_all)(imgs)                    | Performs prediction on the given batch of images.                                                                             |
| [`validate`](#fiftyone.utils.winnow.WinnowModel.validate)(config)                        | Validates that the given config is an instance of <cls>Config.                                                                |

#### *property* has_logits

True; classifications can store their per-option logits.

#### *property* questions

A copy of the question schema answered for each image.

#### *property* ragged_batches

False; [`collate_fn`](#fiftyone.utils.winnow.WinnowModel.collate_fn) pads rows of different lengths.

#### *property* has_collate_fn

True; the model pads its own batches.

#### *property* collate_fn

The function that continues each image’s prompt prefix with every
question and pads the rows.

#### build_get_item(field_mapping=None)

Builds the [`WinnowGetItem`](#fiftyone.utils.winnow.WinnowGetItem) that encodes each image into
this model’s prompt prefix.

* **Parameters:**
  **field_mapping** (*None*) – a dict mapping the GetItem’s required keys
  to field names of the dataset
* **Returns:**
  a [`WinnowGetItem`](#fiftyone.utils.winnow.WinnowGetItem)

#### *property* can_embed_prompts

Whether this model can generate prompt embeddings.

This method returns `False` by default. Models that can generate
prompt embeddings should override this via implementing the
`PromptMixin` interface.

#### *property* classes

The list of class labels for the model, if known.

#### *property* device

The `torch:torch.torch.device` that the model is using.

#### embed(arg)

Generates an embedding for the given data.

Subclasses can override this method to increase efficiency, but, by
default, this method simply calls [`predict()`](#fiftyone.utils.winnow.WinnowModel.predict) and then returns
[`get_embeddings()`](#fiftyone.utils.winnow.WinnowModel.get_embeddings).

* **Parameters:**
  **arg** – the data. See [`predict()`](#fiftyone.utils.winnow.WinnowModel.predict) for details
* **Returns:**
  a numpy array containing the embedding

#### embed_all(args)

Generates embeddings for the given iterable of data.

Subclasses can override this method to increase efficiency, but, by
default, this method simply iterates over the data and applies
[`embed()`](#fiftyone.utils.winnow.WinnowModel.embed) to each.

* **Parameters:**
  **args** – an iterable of data. See [`predict_all()`](#fiftyone.utils.winnow.WinnowModel.predict_all) for details
* **Returns:**
  a numpy array containing the embeddings stacked along axis 0

#### *classmethod* from_config(config)

Instantiates a Configurable class from a <cls>Config instance.

#### *classmethod* from_dict(d)

Instantiates a Configurable class from a <cls>Config dict.

* **Parameters:**
  **d** – a dict to construct a <cls>Config
* **Returns:**
  an instance of cls

#### *classmethod* from_json(json_path)

Instantiates a Configurable class from a <cls>Config JSON file.

* **Parameters:**
  **json_path** – path to a JSON file for type <cls>Config
* **Returns:**
  an instance of cls

#### *classmethod* from_kwargs(\*\*kwargs)

Instantiates a Configurable class from keyword arguments defining
the attributes of a <cls>Config.

* **Parameters:**
  **\*\*kwargs** – keyword arguments that define the fields of a
  <cls>Config dict
* **Returns:**
  an instance of cls

#### get_embeddings()

Returns the embeddings generated by the last forward pass of the
model.

By convention, this method should always return an array whose first
axis represents batch size (which will always be 1 when [`predict()`](#fiftyone.utils.winnow.WinnowModel.predict)
was last used).

* **Returns:**
  a numpy array containing the embedding(s)

#### *property* has_embeddings

Whether this model has embeddings.

#### *property* mask_targets

The mask targets for the model, if any.

#### *property* media_type

The media type processed by the model.

#### *property* num_classes

The number of classes for the model, if known.

#### *static* parse(class_name, module_name=None)

Parses a Configurable subclass name string.

Assumes both the Configurable class and the Config class are defined
in the same module. The module containing the classes will be loaded
if necessary.

* **Parameters:**
  * **class_name** – a string containing the name of the Configurable class,
    e.g. “ClassName”, or a fully-qualified class name, e.g.
    “eta.core.config.ClassName”
  * **module_name** – a string containing the fully-qualified module name,
    e.g. “eta.core.config”, or None if class_name includes the
    module name. Set module_name = \_\_name_\_ to load a class from
    the calling module
* **Returns:**
  the Configurable class
  config_cls: the Config class associated with cls
* **Return type:**
  [cls](fiftyone.brain.internal.core.elasticsearch.md#fiftyone.brain.internal.core.elasticsearch.ElasticsearchSimilarityConfig.cls)

#### predict(img)

Performs prediction on the given image.

* **Parameters:**
  **img** – 

  the image to process, which can be any of the following:
  - A PIL image
  - A uint8 numpy array (HWC)
  - A Torch tensor (CHW)
* **Returns:**
  a [`fiftyone.core.labels.Label`](fiftyone.core.labels.md#fiftyone.core.labels.Label) instance or dict of
  [`fiftyone.core.labels.Label`](fiftyone.core.labels.md#fiftyone.core.labels.Label) instances containing the
  predictions

#### predict_all(imgs)

Performs prediction on the given batch of images.

* **Parameters:**
  **imgs** – 

  the batch of images to process, which can be any of the
  following:
  - A list of PIL images
  - A list of uint8 numpy arrays (HWC)
  - A list of Torch tensors (CHW)
  - A uint8 numpy tensor (NHWC)
  - A Torch tensor (NCHW)
* **Returns:**
  a list of [`fiftyone.core.labels.Label`](fiftyone.core.labels.md#fiftyone.core.labels.Label) instances or a list
  of dicts of [`fiftyone.core.labels.Label`](fiftyone.core.labels.md#fiftyone.core.labels.Label) instances
  containing the predictions

#### *property* preprocess

Whether to apply preprocessing transforms for inference, if any.

#### *property* required_keys

The required keys that must be provided as parameters to methods
like `apply_model()` and `compute_embeddings()` at runtime.

#### *property* skeleton

The keypoint skeleton for the model, if any.

#### *property* store_logits

Whether the model should store logits in its predictions.

#### *property* transforms

A `torchvision.transforms` function that will be applied to each
input before prediction, if any.

#### *property* using_gpu

Whether the model is using GPU.

#### *property* using_half_precision

Whether the model is using half precision.

#### *classmethod* validate(config)

Validates that the given config is an instance of <cls>Config.

* **Raises:**
  **ConfigurableError** – if config is not an instance of <cls>Config