Downloads · 30 days
224
10% of all-time downloads
keyuan01/Gemma-4-26B-A4B-it-MixQ-13G-GGUF
Gemma-4-26B-A4B-it-MixQ-13G-GGUF is a image-text-to-text model from keyuan01. Use it for the image-text-to-text task on the model card, and read the license before you ship it in a product. It is set up for gguf. The card lists the license as apache-2.0.
- Model: googlegemma-4-26b-a4b-it-mix-13GB.gguf - Multimodal projector: mmproj-gemma-4-26b-a4b-it-f16.gguf
Downloads · 30 days
224
10% of all-time downloads
All-time downloads
2.3K
Public
Repo size
14.8 GB
Likes
2
Public
Click a slice to open those files.
.gguf14.8 GB · 100%
From the Hugging Face model README
google_gemma-4-26b-a4b-it-mix-13GB.ggufmmproj-gemma-4-26b-a4b-it-f16.gguf13,638,779,232 bytes12.70 GiB13.64 GB1,193,058,432 bytesThis is the smaller mixed-quant target built from the F16 text GGUF for google/gemma-4-26B-A4B-it.
It is not a pure uniform quant.
It is a mixed recipe built for llama.cpp, with multimodal support preserved through the separate projector file.
This release is a GGUF quantized model for llama.cpp.
Quantization family:
GGUFllama.cppMix-Quant)imatrix-guided quantizationThis is not a single uniform Q4 or Q3 file.
It is a mixed-precision build where different tensor groups keep different quant types according to sensitivity and size budget.
This published 13GB file follows a mixed recipe in this style:
token_embd -> q5_koutput -> q5_krouter -> q8_0attn_q -> q6_kattn_k -> q6_kattn_v -> q6_kattn_output -> q6_kffn_gate_up_exps -> mixed q4_k / q3_kffn_down_exps -> q4_0Notes:
ffn_gate_up_exps is the main mixed expert blockQ6_K + Q4_K/Q3_K expert mix than to a Q3-centered dense recipeimatrix)This release follows the same imatrix-guided quantization idea used in the 31B line.
Core formula:
I_j = Σ_t x_{t,j}^2
Where:
x_{t,j} is the activation value of channel j for token/sample step tI_j is the accumulated importance score of that channel across calibration textPractical meaning:
imatrix does not use benchmark scores directly.
It estimates sensitivity from activations collected on calibration data.
Yes. Multimodal remains supported when paired with:
mmproj-gemma-4-26b-a4b-it-f16.ggufNotes:
mmproj together with the main modelThe working road was:
llama.cpp inference.Observed checks for the published 13GB release:
mmprojNote:
Minimal setup:
llama.cpp with CUDA support.mmproj together if you need vision.Example server:
llama-server \
-m 'google_gemma-4-26b-a4b-it-mix-13GB.gguf' \
--mmproj 'mmproj-gemma-4-26b-a4b-it-f16.gguf' \
-ngl 999 -fa on --ctx-size 4096 -np 1 --port 18081
This repository is a GGUF release of the Google base model.
License:
Apache-2.0https://ai.google.dev/gemma/docs/gemma_4_licenseUse this version if you want:
Gemma 4 26B A4B itmmproj