prithivMLmods/gemma-4-E4B-it-NVFP4
gemma-4-E4B-it-NVFP4 is an NVFP4-compressed evolution of gemma-4-E4B-it. This variant leverages F32 · BF16 · F8E4M3 · U8 precision formats to significantly reduce memory footprint and improve inference efficiency while maintaining strong output quality. gemma-4-E4B-it from Google is a 4.5B effective parameter (8B total with Per-Layer Embeddings) multimodal dense model in the Gemma 4 family, optimi
pip install mlforge-sdk && mlforge pull prithivMLmods/gemma-4-E4B-it-NVFP4
Model details
About prithivMLmods/gemma-4-E4B-it-NVFP4
gemma-4-E4B-it-NVFP4 is an NVFP4-compressed evolution of gemma-4-E4B-it. This variant leverages F32 · BF16 · F8E4M3 · U8 precision formats to significantly reduce memory footprint and improve inference efficiency while maintaining strong output quality. gemma-4-E4B-it from Google is a 4.5B effective parameter (8B total with Per-Layer Embeddings) multimodal dense model in the Gemma 4 family, optimized for edge deployment on laptops, high-end smartphones, and consumer GPUs with native support for text, images (variable aspect ratio and resolution), audio processing, and configurable thinking modes for step-by-step reasoning. Featuring 42 layers, a 512-token sliding window, 128K context length, and a 262K vocabulary, it delivers frontier-level performance in agentic workflows, multilingual OCR and handwriting recognition, document and PDF parsing, UI and screen analysis, chart interpretation, object detection with pointing, coding assistance, and low-latency speech-to-text understanding. Rivaling models 10 to 20× larger while maintaining Google's production-grade safety alignments, the instruction-tuned variant excels at on-device autonomous agents via Android AICore and Qualcomm optimizations, with open weights enabling local-first inference across MediaTek and ARM CPUs as well as NVIDIA RTX GPUs for privacy-focused applications such as mobile IDEs, real-time document processing, and structured data extraction in resource-constrained environments.