G
GLM-5.3 (NVFP4)
Z.ai (NVFP4: Red Hat AI) · GLM
LLMGLM-5.3-LicensepublicTrusted753B params
Ready-made NVFP4 build (compressed-tensors, Red Hat AI) of GLM-5.3 for Blackwell GPUs: 465 GB against 756 GB in FP8.
GLM-5.3 (NVFP4)
Miner
The shard index (model.safetensors.index.json) is 23.3 MB. It needs a miner that takes a shard index up to 64 MiB; the 0.0.10 release refuses any model JSON over 16 MiB.