01

Basic information

Model repository: AMD/granite-4.0-h-small-w4a16-llmcompressor
Organization: AMD
Tasks: text-generation

02

Architecture and runtime

Model type: granitemoehybrid
Architecture: GraniteMoeHybridForCausalLM
Libraries: pytorch, transformer, safetensors
Base models: ibm-granite/granite-4.0-h-small
Current revision: master

03

License and update history

License: apache-2.0
Created: 2026-09-19
Last updated: 2026-09-19

quantizedint4w4a16weight-only4-bitllm-compressorzendnncompressed-tensorsamdcpu-inference
04

Hardware requirements

Not filled in locally.

05

Download links

Not filled in locally.

06

Basic usage example

Not filled in locally.