BulkInferrer
E457350
BulkInferrer is a TensorFlow Extended (TFX) component used to run large-scale batch inference with trained machine learning models on sizable datasets.
Statements (44)
| Predicate | Object |
|---|---|
| instanceOf |
TFX component
ⓘ
batch inference component ⓘ |
| abbreviationOf | TFX BulkInferrer NERFINISHED ⓘ |
| canUse |
TFX Transform outputs
ⓘ
TransformGraph artifact ⓘ |
| category |
MLOps
NERFINISHED
ⓘ
machine learning infrastructure ⓘ |
| compatibleWith |
Apache Airflow
NERFINISHED
ⓘ
Kubeflow Pipelines NERFINISHED ⓘ TFX pipelines NERFINISHED ⓘ Vertex AI Pipelines NERFINISHED ⓘ |
| configuredBy | BulkInferrerSpec NERFINISHED ⓘ |
| designedFor |
production ML workflows
ⓘ
scalable inference ⓘ |
| developedBy | Google NERFINISHED ⓘ |
| documentationUrl | https://www.tensorflow.org/tfx/guide/bulkinferrer ⓘ |
| follows | Trainer component ⓘ |
| hasProperty |
deterministic batch prediction
ⓘ
non-serving, offline inference ⓘ |
| implementedIn | Python NERFINISHED ⓘ |
| inputType | TFRecord files via Examples artifact ⓘ |
| integratesWith |
TFX metadata
NERFINISHED
ⓘ
TFX orchestration NERFINISHED ⓘ |
| license | Apache License 2.0 ⓘ |
| operatesOn |
Examples artifact
ⓘ
TFX Example artifacts ⓘ |
| outputType | TFRecord files with predictions ⓘ |
| partOf | TensorFlow Extended NERFINISHED ⓘ |
| precedes | model evaluation in some pipelines ⓘ |
| produces |
PredictionResults artifact
ⓘ
inference results ⓘ predictions ⓘ |
| requires |
Model artifact
ⓘ
trained model ⓘ |
| softwareLibrary | TensorFlow Extended NERFINISHED ⓘ |
| supports |
SavedModel format
ⓘ
TensorFlow models ⓘ data parallelism ⓘ distributed processing ⓘ |
| usedFor |
generating predictions on sizable datasets
ⓘ
large-scale batch inference ⓘ offline prediction generation ⓘ running inference with trained machine learning models ⓘ scoring examples with a trained model ⓘ |
Referenced by (1)
Full triples — surface form annotated when it differs from this entity's canonical label.