The quantization was performed using the mudler/apex-quant project, focusing on maximizing quality-to-size ratio using importance matrix (imatrix) guided quantization.

📥 Source & Credits

Base Model: ibm-granite/granite-3.1-3b-a800m-instruct
F16 GGUF & Imatrix: The F16 source model and the importance matrix file used for quantization were sourced from bartowski's GGUF repository.

Special thanks to @bartowski for providing the high-quality imatrix file!

⚠️ For technical validation only

Severe accuracy loss due to quantization; outputs may contain hallucinations, gibberish, or fail basic tasks.
Suitable only for researching quantization noise, debugging conversion scripts, or comparing compression artifacts.
No post-training calibration, fine-tuning, or recovery techniques were applied.

Briko/Granite-3.1-3B-A800m-Instruct-APEX

Granite-3.1-3b-a800m-instruct APEX Quantized (GGUF)

📥 Source & Credits

⚠️ For technical validation only

Briko/
Granite-3.1-3B-A800m-Instruct-APEX