The dataset viewer is not available because its heuristics could not detect any supported data files. You can try uploading some data files, or configuring the data files location manually.
Forlinx Embedded: AI Models & Multimedia Resource Hub
Welcome to the official Forlinx Embedded resource repository. This hub serves as a centralized ecosystem providing high-performance Edge AI models, Hardware Development Kits, and Creative Marketing Assets specifically tailored for our SoM (System on Module) and SBC (Single Board Computer) platforms.
Resource Categories
To accelerate your deployment and content creation, our resources are organized into three primary pillars:
| Category | Key Deliverables | Target Platforms |
|---|---|---|
| Edge AI Model Zoo | NPU-optimized weights (Qwen, Llama, YOLO, MobileNet) | RK3588, i.MX95, RK3576, etc. |
| Development Resources | Pre-built OS images, NPU toolchains (RKNN/eIQ), and SDKs | Forlinx SoM Boards |
| Creator Toolbox | 4K product renders, application demos, and KOL B-roll footage | Marketers, Tech Bloggers, KOLs |
1. Hardware-Specific Ecosystems
🚀 Rockchip RK3588 Series
- NPU Capabilities: 6 TOPS triple-core NPU support.
- Resources: Optimized
.rknnmodels, 8K hardware encoding/decoding samples, and multi-channel ISP configurations for high-end vision tasks.
🛡️ NXP i.MX95 Series
- NPU Capabilities: Integrated eIQ Neutron Neural Processing Unit.
- Resources: Industrial-grade safety demos, real-time control + AI fusion frameworks, and optimized models for machine vision and human-machine interface (HMI).
2. Optimized Model Zoo (Edge AI)
We bridge the gap between heavy cloud models and resource-constrained edge devices:
- Large Language Models (LLM): Qwen-1.5B/7B and Llama-3 (INT4/INT8 quantized) optimized for low-latency terminal interaction.
- Computer Vision: YOLOv8, YOLOv10, and Segment Anything (SAM) tuned for real-time inference on Forlinx hardware.
- Multi-modal: Speech-to-Text (Whisper) and Image Captioning models for interactive IoT devices.
3. Creative Media & KOL Support
This section is dedicated to our marketing partners and content creators.
- High-Res Assets: Transparent PNGs of Forlinx modules and 3D exploded view animations.
- Application Scenarios: Raw video footage of smart medical, industrial automation, and smart city deployments for use in technical reviews.
- Brand Guidelines: Official Forlinx logos, typography, and presentation templates for global consistency.
How to Utilize These Resources
For Developers
- Clone/Download: Use
git lfs cloneor the Hugging Face CLI to fetch optimized model weights. - Deployment: Refer to the
deployment_guide.mdin each subdirectory for instructions on flashing images and running NPU inference.
For Content Creators
Navigate to the /assets or /media directory to find high-quality visuals. Please ensure the use of these materials aligns with the Forlinx Brand Guidelines.
Technical Constraints & Risks
- Hardware Dependency: AI models provided here (especially
.rknnor.nbformats) are strictly compiled for specific NPU architectures. - Environmental Factors: Performance metrics (Inference speed/FPS) may vary based on thermal management and power supply configurations of the carrier board.
About Forlinx Embedded
Forlinx Embedded is a global leader in the design and manufacture of high-performance embedded solutions. As an official partner of Rockchip, NXP, TI, and Allwinner, we provide "Platform + Ecosystem" support to simplify the path from prototype to production.
- Global Website: www.forlinx.net
- Technical Support: support@forlinx.com
- Sales Team: sales@forlinx.com
Licensing
- Code & Models: Distributed under the Apache 2.0 License unless otherwise specified.
- Media Assets: Distributed under CC BY-NC-ND 4.0 (Attribution-NonCommercial-NoDerivs).
- Downloads last month
- 1,607