`sparknlp.annotator.cv.smolvlm_transformer`#

Module Contents#

Classes#

SmolVLMTransformer

SmolVLMTransformer can load SmolVLM models for visual question answering. The model

class SmolVLMTransformer(classname='com.johnsnowlabs.nlp.annotators.cv.SmolVLMTransformer', java_model=None)[source]#

SmolVLMTransformer can load SmolVLM models for visual question answering. The model consists of a vision encoder, a text encoder as well as a text decoder. The vision encoder will encode the input image, the text encoder will encode the input question together with the encoding of the image, and the text decoder will output the answer to the question.

SmolVLM is a compact open multimodal model that accepts arbitrary sequences of image and text inputs to produce text outputs. Designed for efficiency, SmolVLM can answer questions about images, describe visual content, create stories grounded on multiple images, or function as a pure language model without visual inputs. Its lightweight architecture makes it suitable for on-device applications while maintaining strong performance on multimodal tasks.

Pretrained models can be loaded with pretrained() of the companion object: >>> visualQA = SmolVLMTransformer.pretrained() … .setInputCols([“image_assembler”]) … .setOutputCol(“answer”)

The default model is “smolvlm_instruct_int4”, if no name is provided. For available pretrained models, refer to the Models Hub.

Models from the HuggingFace 🧧 Transformers library are also compatible with Spark NLP 🚀. To check compatibility and learn how to import them, see Import Transformers into Spark NLP 🚀. For extended examples, refer to the SmolVLMTransformer Test Suite.

Input Annotation types	Output Annotation type
`IMAGE`	`DOCUMENT`

Parameters:

batchSizeint, optional: Batch size. Larger values allow faster processing but require more memory, by default 1.
configProtoBytesbytes, optional: ConfigProto from TensorFlow, serialized into a byte array.
maxSentenceLengthint, optional: Maximum sentence length to process, by default 20.
doImageSplittingbool, optional: Whether to split the image, by default True.
imageTokenint, optional: Token ID for image embeddings, by default 49153.
numVisionTokensint, optional: Number of vision tokens, by default 81.
maxImageSizeint, optional: Maximum image size for the model, by default 384.
patchSizeint, optional: Patch size for the model, by default 14.
paddingConstantint, optional: Padding constant for the model, by default 0.

Examples

>>> import sparknlp
>>> from sparknlp.base import *
>>> from sparknlp.annotator import *
>>> from pyspark.ml import Pipeline
>>> from pyspark.sql.functions import lit
>>> imageDF = spark.read.format("image").load(path=images_path)
>>> testDF = imageDF.withColumn(
...     "text",
...     lit("<|im_start|>User:<image>Can you describe the image?<end_of_utterance>\nAssistant:")
... )
>>> imageAssembler = ImageAssembler() \
...     .setInputCol("image") \
...     .setOutputCol("image_assembler")
>>> visualQAClassifier = SmolVLMTransformer.pretrained() \
...     .setInputCols("image_assembler") \
...     .setOutputCol("answer")
>>> pipeline = Pipeline().setStages([
...     imageAssembler,
...     visualQAClassifier
... ])
>>> result = pipeline.fit(testDF).transform(testDF)
>>> result.select("image_assembler.origin", "answer.result").show(truncate=False)
+--------------------------------------+----------------------------------------------------------------------+
|origin                                |result                                                                |
+--------------------------------------+----------------------------------------------------------------------+
|[file:///content/images/cat_image.jpg]|[The unusual aspect of this picture is the presence of two cats lying on a pink couch]|
+--------------------------------------+----------------------------------------------------------------------+

name = 'SmolVLMTransformer'[source]#

inputAnnotatorTypes[source]#

outputAnnotatorType = 'document'[source]#

minOutputLength[source]#

maxOutputLength[source]#

doSample[source]#

temperature[source]#

topK[source]#

topP[source]#

repetitionPenalty[source]#

noRepeatNgramSize[source]#

ignoreTokenIds[source]#

beamSize[source]#

stopTokenIds[source]#

imageToken[source]#

numVisionTokens[source]#

maxImageSize[source]#

patchSize[source]#

paddingConstant[source]#

doImageSplitting[source]#

setMaxSentenceSize(value)[source]#: Sets Maximum sentence length that the annotator will process, by default 20. Parameters ———- value : int

Maximum sentence length that the annotator will process

setIgnoreTokenIds(value)[source]#: A list of token ids which are ignored in the decoder’s output. Parameters ———- value : List[int]

The words to be filtered out

setStopTokenIds(value)[source]#: Stop tokens to terminate the generation. Parameters ———- value : List[int]

The tokens that terminate generation

setConfigProtoBytes(b)[source]#: Sets configProto from tensorflow, serialized into byte array. Parameters ———- b : List[int]

ConfigProto from tensorflow, serialized into byte array

setMinOutputLength(value)[source]#: Sets minimum length of the sequence to be generated. Parameters ———- value : int

Minimum length of the sequence to be generated

setMaxOutputLength(value)[source]#: Sets maximum length of output text. Parameters ———- value : int

Maximum length of output text

setDoSample(value)[source]#: Sets whether or not to use sampling, use greedy decoding otherwise. Parameters ———- value : bool

Whether or not to use sampling; use greedy decoding otherwise

setTemperature(value)[source]#: Sets the value used to module the next token probabilities. Parameters ———- value : float

The value used to module the next token probabilities

setTopK(value)[source]#: Sets the number of highest probability vocabulary tokens to keep for top-k-filtering. Parameters ———- value : int

Number of highest probability vocabulary tokens to keep

setTopP(value)[source]#: Sets the top cumulative probability for vocabulary tokens. If set to float < 1, only the most probable tokens with probabilities that add up to topP or higher are kept for generation. Parameters ———- value : float

Cumulative probability for vocabulary tokens

setRepetitionPenalty(value)[source]#: Sets the parameter for repetition penalty. 1.0 means no penalty. Parameters ———- value : float

The repetition penalty

References#

See Ctrl: A Conditional Transformer Language Model For Controllable Generation for more details.

setNoRepeatNgramSize(value)[source]#: Sets size of n-grams that can only occur once. If set to int > 0, all ngrams of that size can only occur once. Parameters ———- value : int

N-gram size can only occur once

setBeamSize(value)[source]#: Sets the number of beam size for beam search, by default 1. Parameters ———- value : int

Number of beam size for beam search

setImageToken(value)[source]#: Sets the token ID for image embeddings. Parameters ———- value : int

Token ID for image embeddings

setNumVisionTokens(value)[source]#: Sets the number of vision tokens. Parameters ———- value : int

Number of vision tokens

setMaxImageSize(value)[source]#: Sets the maximum image size for the model. Parameters ———- value : int

Maximum image size

setPatchSize(value)[source]#: Sets the patch size for the model. Parameters ———- value : int

Patch size

setPaddingConstant(value)[source]#: Sets the padding constant for the model. Parameters ———- value : int

Padding constant

setDoImageSplitting(value)[source]#: Sets whether to split the image. Parameters ———- value : bool

Whether to split the image

static loadSavedModel(folder, spark_session, use_openvino=False)[source]#

Loads a locally saved model. Parameters ———- folder : str

Folder of the saved model

spark_sessionpyspark.sql.SparkSession: The current SparkSession
use_openvinobool, optional: Whether to use OpenVINO for inference, by default False

Returns#

SmolVLMTransformer: The restored model

static pretrained(name='smolvlm_instruct_int4', lang='en', remote_loc=None)[source]#

Downloads and loads a pretrained model. Parameters ———- name : str, optional

Name of the pretrained model, by default “smolvlm_instruct_int4”

langstr, optional: Language of the pretrained model, by default “en”
remote_locstr, optional: Optional remote address of the resource, by default None. Will use Spark NLPs repositories otherwise.

Returns#

SmolVLMTransformer: The restored model

sparknlp.annotator.cv.smolvlm_transformer#

Module Contents#

Classes#

References#

Returns#

Returns#

`sparknlp.annotator.cv.smolvlm_transformer`#