GPU-Enabled TensorFlow Deployment
A pre-trained TensorFlow model is a machine learning model that has already been trained on a dataset and whose learned weights can be reused for inference or further fine-tuning. In SAP AI Core, developers use pre-trained TensorFlow models together with GPU resources to accelerate workloads such as image recognition or other deep learning tasks without training a model from scratch.
Tutorials that teach this
Prerequisites
- Concept Pre-trained Model Integration The concept name was not provided, and the available source snippets do not contain enough shared or specific detail to identify and define a single, unambiguous SAP developer concept. Please provide the concept name or additional source content so an accurate, grounded definition can be written.
- Concept SAP AI Core Service Setup The concept name was not provided, so a precise definition cannot be constructed from the available sources. Based on the sources, these materials collectively cover building and integrating [Generative AI applications](https://architecture.learning.sap.com/docs/golden-path/ai-golden-path/2-build-and-deliver/3-genai-applications/readme) on SAP Business Technology Platform using SAP AI Core — a managed service that developers use to train models, run AI workloads, and connect to large language models via orchestration. Developers leverage it alongside [SAP BTP service instances and keys](https://help.sap.com/docs/btp/sap-business-technology-platform/5b35ee98403045309bb0f8c7f0c79365?locale=en-US&state=PRODUCTION&version=Cloud) to authenticate and consume AI capabilities such as prompt optimization, grounding, multimodal responses, and custom evaluation within their applications.
- Concept SAP AI Core Model Deployment and Inference The concept name provided is "undefined," and the supplied sources do not contain a definition or description of a specific, identifiable SAP developer concept that can be documented. Without a valid concept name or sufficient source content to ground a definition, a accurate and source-supported definition cannot be produced.
- Concept Docker Image Build and Publish The provided sources do not contain sufficient information to write a grounded definition for an **undefined** concept. The sources cover a range of SAP BTP topics (HANA connection issues, Kyma runtime, AI Core, Cloud Foundry) but do not define or describe a specific named concept that can be documented here.
- Concept Cloud Object Store Integration for ML Data The provided sources do not contain enough information to define a specific SAP developer concept, as no concept name was supplied and the source snippets consist only of tutorial titles without descriptive content. A meaningful, source-grounded definition cannot be written without at least a named concept or substantive excerpt from the referenced materials.