This is an FP8-dynamic quantized version of the 31-billion parameter Gemma 4 instruction-tuned (it) model, published by RedHatAI on Hugging Face. It enables efficient inference of a large language model while preserving most of the original performance. The model includes a chat template and is designed for developers seeking optimized open-weight LLMs.
Gemma 4 31B It FP8 Dynamic sits in PulseGate's Foundation models & chat category. It focuses on running large instruction-tuned language models efficiently on hardware with reduced memory and compute requirements. It is built as an open-source project for developers. Gemma 4 31B It FP8 Dynamic is open source under the Apache-2.0 license. The product ships for the web, the command line, and API.
It is developed by RedHatAI, and the product first shipped in 2019. Development happens publicly on GitHub with 3.6k stars and 171 commits in the last 90 days. Key capabilities include instruction tuned, FP8 quantization, and dynamic quantization.
Latest indexed changes and source events
RedHatAI/gemma-4-31B-it-FP8-dynamic verified by the PulseGate indexer
