>> zerofata's GLM-4.5-Iceblink-v2-106B-A12B-exl3

[BASE :: GLM-4.5-AIR]

[TUNE :: ZEROFATA]

[QUANT :: UNSTABLELLAMA]

// REPO

EXL3 quantizations of zerofata's ICEBLINK-v2.

Quantized with exllamav3 0.0.12.

[BRANCH]	[GiB]	[K/L_DIV]	[PPL]
2.05bpw	26.8	0.883	5.676
2.5bpw	32.3	0.591	5.261
3.05bpw	39.1	0.199	4.513
4.0bpw	51	0.069	4.289
5.0bpw	63.3	0.026	4.183
bf16	205.8	0	4.132

Use HF-CLI to pull specific branches to your local machine:

huggingface-cli download UnstableLlama/zerofata_GLM-4.5-Iceblink-v2-106B-A12B-exl3 --revision "3.05bpw" --local-dir ./

Downloads last month: -; Downloads are not tracked for this model. How to track

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Base model

Finetuned

Quantized

(7)

this model