gemma-4-26B-A4B-TQ3_1S-GGUF
TQ3_1S (ternary) quantization of Google's Gemma 4 26B-A4B MoE model, produced
with llama.cpp.
Quant Details
| Quant | Type |
|---|
| TQ3_1S | Ternary quantization, 1-bit scale — aggressive size reduction with reasonable quality retention on MoE architectures |
Usage
Load with a llama.cpp build that supports TQ3_1S (recent builds required).
Original Model
