1 link tagged with all of: local inference + quantization + dynamic gguf + llm + unsloth
Click any tag below to further narrow down your results
Links
This guide shows how to run Z.ai’s open-source GLM-5.2 model on local hardware using Unsloth Dynamic GGUF quantizations. It covers memory requirements for 1-bit and 2-bit setups, recommended inference settings, and step-by-step instructions for Unsloth Studio and llama.cpp. The article also explains KLD benchmarks and quantization accuracy trade-offs.
- GLM-5.2's 744B-parameter model can run on a 24GB GPU plus 256GB RAM using Unsloth's 2-bit dynamic quant (239GB, ~82% top-1 accuracy), an 84% size cut from the full 1.5TB model.
- The 1-bit quant shrinks total memory needs to ~223GB but drops accuracy to ~76.2%.
- 4-bit and 5-bit dynamic quants are nearly lossless per KLD testing, making them the better choice when accuracy on out-of-distribution tasks matters most.
- Dynamic quantization keeps critical layers at higher precision while aggressively compressing less important ones, which is why even 1-2 bit versions stay broadly usable despite the accuracy trade-off.