> For the complete documentation index, see [llms.txt](https://deeplearning4j.konduit.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://deeplearning4j.konduit.ai/model-import.md). # Model Import - [Overview](https://deeplearning4j.konduit.ai/model-import/overview.md): All model import paths in DL4J — Keras to DL4J, TF/ONNX to SameDiff, and direct inference runtimes - [Keras Import](https://deeplearning4j.konduit.ai/model-import/overview-1.md): Importing Keras models into Deeplearning4j — supported features, limitations, and getting started - [Getting Started](https://deeplearning4j.konduit.ai/model-import/overview-1/getting-started.md): Step-by-step guide to importing Keras models — saving in Python, loading in Java, and running inference - [Functional Model](https://deeplearning4j.konduit.ai/model-import/overview-1/functional-model.md): Importing Keras Functional API models as ComputationGraph - [Sequential Model](https://deeplearning4j.konduit.ai/model-import/overview-1/sequential-model.md): Importing Keras Sequential models as MultiLayerNetwork - [API Reference](https://deeplearning4j.konduit.ai/model-import/overview-1/model-import-api.md): KerasModelImport API — all import methods and configuration options - [Supported Features](https://deeplearning4j.konduit.ai/model-import/overview-1/supported-features.md): Full support matrix for Keras model import — layers, activations, losses, and optimizers - [Core Layers](https://deeplearning4j.konduit.ai/model-import/overview-1/layers-core.md): DL4J equivalents and API reference for Keras core layers — Dense, Flatten, Dropout, Reshape, Merge, Permute, and more. - [Convolutional Layers](https://deeplearning4j.konduit.ai/model-import/overview-1/layers-convolutional.md): DL4J equivalents and API reference for Keras convolutional layers — Conv1D, Conv2D, Conv3D, SeparableConv2D, transposed convolutions, cropping, upsampling, and zero-padding. - [Recurrent Layers](https://deeplearning4j.konduit.ai/model-import/overview-1/layers-recurrent.md): DL4J equivalents and API reference for Keras recurrent layers — SimpleRNN, LSTM, and associated utilities. - [Pooling Layers](https://deeplearning4j.konduit.ai/model-import/overview-1/layers-pooling.md): DL4J equivalents and API reference for Keras pooling layers — MaxPooling, AveragePooling, and GlobalPooling in 1D, 2D, and 3D variants. - [Normalization Layers](https://deeplearning4j.konduit.ai/model-import/overview-1/layers-normalization.md): DL4J equivalents and API reference for Keras normalization layers — BatchNormalization. - [Embedding Layers](https://deeplearning4j.konduit.ai/model-import/overview-1/layers-embeddings.md): DL4J equivalents and API reference for Keras embedding layers — Embedding mapped to EmbeddingSequenceLayer. - [Advanced Activations](https://deeplearning4j.konduit.ai/model-import/overview-1/layers-advanced-activations.md): DL4J equivalents and API reference for Keras advanced activation layers — LeakyReLU, PReLU, ELU, and ThresholdedReLU. - [Wrapper Layers](https://deeplearning4j.konduit.ai/model-import/overview-1/layers-wrappers.md): DL4J equivalents and API reference for Keras wrapper layers — Bidirectional and TimeDistributed. - [Noise Layers](https://deeplearning4j.konduit.ai/model-import/overview-1/layers-noise.md): DL4J equivalents and API reference for Keras noise layers — GaussianNoise, GaussianDropout, and AlphaDropout. - [Local Layers](https://deeplearning4j.konduit.ai/model-import/overview-1/layers-local.md): DL4J equivalents and API reference for Keras locally connected layers — LocallyConnected1D and LocallyConnected2D. - [Activations](https://deeplearning4j.konduit.ai/model-import/overview-1/activations.md): Keras to DL4J activation function mapping for model import - [Optimizers](https://deeplearning4j.konduit.ai/model-import/overview-1/optimizers.md): Keras to DL4J optimizer mapping for model import - [Losses](https://deeplearning4j.konduit.ai/model-import/overview-1/losses.md): Keras to DL4J loss function mapping for model import - [Initializers](https://deeplearning4j.konduit.ai/model-import/overview-1/initializers.md): Mapping of Keras weight initializers to DL4J WeightInit implementations for model import in Eclipse Deeplearning4j 1.0.0-M2.1. - [Constraints](https://deeplearning4j.konduit.ai/model-import/overview-1/constraints.md): Mapping of Keras weight constraints to DL4J LayerConstraint implementations for model import in Eclipse Deeplearning4j 1.0.0-M2.1. - [Regularizers](https://deeplearning4j.konduit.ai/model-import/overview-1/regularizers.md): Mapping of Keras regularizers to DL4J regularization for model import in Eclipse Deeplearning4j 1.0.0-M2.1. - [Backend](https://deeplearning4j.konduit.ai/model-import/overview-1/backend.md): DL4J Keras model import is backend-agnostic — models trained with TensorFlow, Theano, or CNTK backends can all be imported into Eclipse Deeplearning4j 1.0.0-M2.1. - [SameDiff Import](https://deeplearning4j.konduit.ai/model-import/overview-2.md): Importing TensorFlow and ONNX models into SameDiff — architecture, supported ops, and usage - [TensorFlow Import](https://deeplearning4j.konduit.ai/model-import/overview-2/tensorflow.md): Importing TensorFlow frozen graphs and SavedModels into SameDiff - [ONNX Import](https://deeplearning4j.konduit.ai/model-import/overview-2/onnx.md): Importing ONNX models into SameDiff - [ONNX Import & Export (Expanded)](https://deeplearning4j.konduit.ai/model-import/overview-2/onnx-expanded.md): ~120 new ONNX op implementations including Microsoft LLM contrib ops, ONNX ML domain classifiers, quantized inference ops, and bidirectional SameDiff-to-ONNX export - [GGML/GGUF Import](https://deeplearning4j.konduit.ai/model-import/overview-3.md): Import quantized LLMs from GGUF files — architecture handlers, quantization codecs, adaptive quantization, round-trip export, and pipeline modules - [ONNX Runtime](https://deeplearning4j.konduit.ai/model-import/overview-4.md): Direct ONNX model inference via ONNX Runtime 1.10 — no conversion to SameDiff required - [TensorFlow](https://deeplearning4j.konduit.ai/model-import/overview-5.md): Running TensorFlow frozen graphs directly via JavaCPP TF bindings - [TensorFlow Lite](https://deeplearning4j.konduit.ai/model-import/overview-5/tensorflow-lite.md): TensorFlow Lite 2.8 inference for mobile and edge deployment - [Apache TVM](https://deeplearning4j.konduit.ai/model-import/overview-6.md): Apache TVM 0.8 integration for optimized model inference --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://deeplearning4j.konduit.ai/model-import.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.