Source code for sparknlp.annotator.cv.smolvlm_transformer
# Copyright 2017-2024 John Snow Labs
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
from sparknlp.common import *
[docs]class SmolVLMTransformer(AnnotatorModel,
HasBatchedAnnotateImage,
HasImageFeatureProperties,
HasEngine,
HasCandidateLabelsProperties,
HasRescaleFactor):
"""
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 :meth:`.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
<https://sparknlp.org/models?task=Question+Answering>`__.
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 🚀
<https://github.com/JohnSnowLabs/spark-nlp/discussions/5669>`_.
For extended examples, refer to the `SmolVLMTransformer Test Suite
<https://github.com/JohnSnowLabs/spark-nlp/blob/master/src/test/scala/com/johnsnowlabs/nlp/annotators/cv/SmolVLMTransformerTest.scala>`_.
====================== ======================
Input Annotation types Output Annotation type
====================== ======================
``IMAGE`` ``DOCUMENT``
====================== ======================
Parameters
----------
batchSize : int, optional
Batch size. Larger values allow faster processing but require more memory,
by default 1.
configProtoBytes : bytes, optional
ConfigProto from TensorFlow, serialized into a byte array.
maxSentenceLength : int, optional
Maximum sentence length to process, by default 20.
doImageSplitting : bool, optional
Whether to split the image, by default True.
imageToken : int, optional
Token ID for image embeddings, by default 49153.
numVisionTokens : int, optional
Number of vision tokens, by default 81.
maxImageSize : int, optional
Maximum image size for the model, by default 384.
patchSize : int, optional
Patch size for the model, by default 14.
paddingConstant : int, 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"
inputAnnotatorTypes = [AnnotatorType.IMAGE]
outputAnnotatorType = AnnotatorType.DOCUMENT
minOutputLength = Param(Params._dummy(), "minOutputLength", "Minimum length of the sequence to be generated",
typeConverter=TypeConverters.toInt)
maxOutputLength = Param(Params._dummy(), "maxOutputLength", "Maximum length of output text",
typeConverter=TypeConverters.toInt)
doSample = Param(Params._dummy(), "doSample", "Whether or not to use sampling; use greedy decoding otherwise",
typeConverter=TypeConverters.toBoolean)
temperature = Param(Params._dummy(), "temperature", "The value used to module the next token probabilities",
typeConverter=TypeConverters.toFloat)
topK = Param(Params._dummy(), "topK",
"The number of highest probability vocabulary tokens to keep for top-k-filtering",
typeConverter=TypeConverters.toInt)
topP = Param(Params._dummy(), "topP",
"If set to float < 1, only the most probable tokens with probabilities that add up to ``top_p`` or higher are kept for generation",
typeConverter=TypeConverters.toFloat)
repetitionPenalty = Param(Params._dummy(), "repetitionPenalty",
"The parameter for repetition penalty. 1.0 means no penalty. See `this paper <https://arxiv.org/pdf/1909.05858.pdf>`__ for more details",
typeConverter=TypeConverters.toFloat)
noRepeatNgramSize = Param(Params._dummy(), "noRepeatNgramSize",
"If set to int > 0, all ngrams of that size can only occur once",
typeConverter=TypeConverters.toInt)
ignoreTokenIds = Param(Params._dummy(), "ignoreTokenIds",
"A list of token ids which are ignored in the decoder's output",
typeConverter=TypeConverters.toListInt)
beamSize = Param(Params._dummy(), "beamSize",
"The Number of beams for beam search.",
typeConverter=TypeConverters.toInt)
stopTokenIds = Param(Params._dummy(), "stopTokenIds",
"Stop tokens to terminate the generation",
typeConverter=TypeConverters.toListInt)
imageToken = Param(Params._dummy(), "imageToken",
"Token id for image embeddings",
typeConverter=TypeConverters.toInt)
numVisionTokens = Param(Params._dummy(), "numVisionTokens",
"Number of vision tokens",
typeConverter=TypeConverters.toInt)
maxImageSize = Param(Params._dummy(), "maxImageSize",
"Maximum image size for the model",
typeConverter=TypeConverters.toInt)
patchSize = Param(Params._dummy(), "patchSize",
"Patch size for the model",
typeConverter=TypeConverters.toInt)
paddingConstant = Param(Params._dummy(), "paddingConstant",
"Padding constant for the model",
typeConverter=TypeConverters.toInt)
doImageSplitting = Param(Params._dummy(), "doImageSplitting",
"Whether to split the image",
typeConverter=TypeConverters.toBoolean)
[docs] def setMaxSentenceSize(self, value):
"""Sets Maximum sentence length that the annotator will process, by
default 20.
Parameters
----------
value : int
Maximum sentence length that the annotator will process
"""
return self._set(maxSentenceLength=value)
[docs] def setIgnoreTokenIds(self, value):
"""A list of token ids which are ignored in the decoder's output.
Parameters
----------
value : List[int]
The words to be filtered out
"""
return self._set(ignoreTokenIds=value)
[docs] def setStopTokenIds(self, value):
"""Stop tokens to terminate the generation.
Parameters
----------
value : List[int]
The tokens that terminate generation
"""
return self._set(stopTokenIds=value)
[docs] def setConfigProtoBytes(self, b):
"""Sets configProto from tensorflow, serialized into byte array.
Parameters
----------
b : List[int]
ConfigProto from tensorflow, serialized into byte array
"""
return self._set(configProtoBytes=b)
[docs] def setMinOutputLength(self, value):
"""Sets minimum length of the sequence to be generated.
Parameters
----------
value : int
Minimum length of the sequence to be generated
"""
return self._set(minOutputLength=value)
[docs] def setMaxOutputLength(self, value):
"""Sets maximum length of output text.
Parameters
----------
value : int
Maximum length of output text
"""
return self._set(maxOutputLength=value)
[docs] def setDoSample(self, value):
"""Sets whether or not to use sampling, use greedy decoding otherwise.
Parameters
----------
value : bool
Whether or not to use sampling; use greedy decoding otherwise
"""
return self._set(doSample=value)
[docs] def setTemperature(self, value):
"""Sets the value used to module the next token probabilities.
Parameters
----------
value : float
The value used to module the next token probabilities
"""
return self._set(temperature=value)
[docs] def setTopK(self, value):
"""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
"""
return self._set(topK=value)
[docs] def setTopP(self, value):
"""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
"""
return self._set(topP=value)
[docs] def setRepetitionPenalty(self, value):
"""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 <https://arxiv.org/pdf/1909.05858.pdf>`__ for more details.
"""
return self._set(repetitionPenalty=value)
[docs] def setNoRepeatNgramSize(self, value):
"""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
"""
return self._set(noRepeatNgramSize=value)
[docs] def setBeamSize(self, value):
"""Sets the number of beam size for beam search, by default `1`.
Parameters
----------
value : int
Number of beam size for beam search
"""
return self._set(beamSize=value)
[docs] def setImageToken(self, value):
"""Sets the token ID for image embeddings.
Parameters
----------
value : int
Token ID for image embeddings
"""
return self._set(imageToken=value)
[docs] def setNumVisionTokens(self, value):
"""Sets the number of vision tokens.
Parameters
----------
value : int
Number of vision tokens
"""
return self._set(numVisionTokens=value)
[docs] def setMaxImageSize(self, value):
"""Sets the maximum image size for the model.
Parameters
----------
value : int
Maximum image size
"""
return self._set(maxImageSize=value)
[docs] def setPatchSize(self, value):
"""Sets the patch size for the model.
Parameters
----------
value : int
Patch size
"""
return self._set(patchSize=value)
[docs] def setPaddingConstant(self, value):
"""Sets the padding constant for the model.
Parameters
----------
value : int
Padding constant
"""
return self._set(paddingConstant=value)
[docs] def setDoImageSplitting(self, value):
"""Sets whether to split the image.
Parameters
----------
value : bool
Whether to split the image
"""
return self._set(doImageSplitting=value)
@keyword_only
def __init__(self, classname="com.johnsnowlabs.nlp.annotators.cv.SmolVLMTransformer",
java_model=None):
super(SmolVLMTransformer, self).__init__(
classname=classname,
java_model=java_model
)
self._setDefault(
batchSize=1,
minOutputLength=0,
maxOutputLength=20,
doSample=False,
temperature=0.6,
topK=-1,
topP=0.9,
repetitionPenalty=1.0,
noRepeatNgramSize=3,
ignoreTokenIds=[],
beamSize=1,
stopTokenIds=[49154],
imageToken=49153,
numVisionTokens=81,
maxImageSize=384,
patchSize=14,
paddingConstant=0,
doImageSplitting=True
)
[docs] @staticmethod
def loadSavedModel(folder, spark_session, use_openvino=False):
"""Loads a locally saved model.
Parameters
----------
folder : str
Folder of the saved model
spark_session : pyspark.sql.SparkSession
The current SparkSession
use_openvino : bool, optional
Whether to use OpenVINO for inference, by default False
Returns
-------
SmolVLMTransformer
The restored model
"""
from sparknlp.internal import _SmolVLMTransformerLoader
jModel = _SmolVLMTransformerLoader(folder, spark_session._jsparkSession, use_openvino)._java_obj
return SmolVLMTransformer(java_model=jModel)
[docs] @staticmethod
def pretrained(name="smolvlm_instruct_int4", lang="en", remote_loc=None):
"""Downloads and loads a pretrained model.
Parameters
----------
name : str, optional
Name of the pretrained model, by default
"smolvlm_instruct_int4"
lang : str, optional
Language of the pretrained model, by default "en"
remote_loc : str, optional
Optional remote address of the resource, by default None. Will use
Spark NLPs repositories otherwise.
Returns
-------
SmolVLMTransformer
The restored model
"""
from sparknlp.pretrained import ResourceDownloader
return ResourceDownloader.downloadModel(SmolVLMTransformer, name, lang, remote_loc)