LLM EXPLORER 60,587 MODELS INDEXED

GLM 5.3 Quark MXFP4 AttnFP8 by amd

By amd · 1317 downloads
HF Repo ↗ Copy URL

GLM 5.3 Quark MXFP4 AttnFP8 is an open-source language model by amd. Features: 384.3b LLM, VRAM: 126.7GB, Context: 1024K, License: other, MoE, LLM Explorer Score: 0.33.

  8-bit Base model:quantized:zai-org/g...   Base model:zai-org/glm-5.3   Conversational   En   Endpoints compatible   Eval-results   Glm moe dsa   Moe   Quark   Region:us   Safetensors   Sharded   Tensorflow   Zh
Model Card on HF 🤗: https://huggingface.co/amd/GLM-5.3-Quark-MXFP4-AttnFP8 

GLM 5.3 Quark MXFP4 AttnFP8 Parameters and Internals

LLM NameGLM 5.3 Quark MXFP4 AttnFP8
Repository 🤗https://huggingface.co/amd/GLM-5.3-Quark-MXFP4-AttnFP8 
Base Model(s)  zai-org/GLM-5.3   zai-org/GLM-5.3
Model Size384.3b
Required VRAM126.7 GB
Updated2026-09-06
Maintaineramd
Model Typeglm_moe_dsa
Model Files  4.8 GB: 1-of-141   2.9 GB: 2-of-141   2.9 GB: 3-of-141   2.9 GB: 4-of-141   2.9 GB: 5-of-141   2.9 GB: 6-of-141   2.9 GB: 7-of-141   2.9 GB: 8-of-141   2.9 GB: 9-of-141   2.9 GB: 10-of-141   2.9 GB: 11-of-141   2.9 GB: 12-of-141   2.9 GB: 13-of-141   2.9 GB: 14-of-141   2.9 GB: 15-of-141   2.9 GB: 16-of-141   2.9 GB: 17-of-141   2.9 GB: 18-of-141   2.9 GB: 19-of-141   3.0 GB: 20-of-141   2.9 GB: 21-of-141   2.9 GB: 22-of-141   2.9 GB: 23-of-141   2.9 GB: 24-of-141   2.9 GB: 25-of-141   2.9 GB: 26-of-141   2.9 GB: 27-of-141   2.9 GB: 28-of-141   2.9 GB: 29-of-141   2.9 GB: 30-of-141   2.9 GB: 31-of-141   2.9 GB: 32-of-141   2.9 GB: 33-of-141   2.9 GB: 34-of-141   2.9 GB: 35-of-141   2.9 GB: 36-of-141   2.9 GB: 37-of-141   2.9 GB: 38-of-141   2.9 GB: 39-of-141   2.9 GB: 40-of-141   2.9 GB: 41-of-141   2.9 GB: 42-of-141   2.9 GB: 43-of-141
Supported Languagesen zh
Model ArchitectureGlmMoeDsaForCausalLM
Licenseother
Context Length1048576
Model Max Length1048576
Transformers Version5.15.0
Vocabulary Size154880