Quantization format
TQ1_0
TQ1_0 has a nominal rate of — bits per weight. We have no measured files carrying this label yet.
From the file· 14 files measured
Nominal bpw
—
from the block layout
Measured average
3.256
14 files
Range
1.811–10.411
varies by architecture
File sizes
0.02 GiB+
up to 92.59 GiB
Real files
one per model, most downloaded first
| Model | Size● | Effective bpw● | vs nominal● |
|---|---|---|---|
| Qwen-Image-Edit-2509 | 6.50 GiB | 2.731 | — |
| Qwen3-0.6B-Base | 0.29 GiB | 4.218 | — |
| Llama-4-Maverick-17B-128E-Instruct | 92.59 GiB | 1.981 | — |
| Wan_2.2_ComfyUI_Repackaged | 1.31 GiB | — | — |
| gemma-4-26B-A4B | 7.98 GiB | 2.583 | — |
| Llama-3.2-3B | 0.86 GiB | 2.307 | — |
| diffusiongemma-26B-A4B-it-HERETIC-Uncensored | 7.99 GiB | 2.657 | — |
| Qwen3.5-9B-Base | 2.68 GiB | 2.388 | — |
| bitnet-b1.58-2B-4T | 1.03 GiB | 10.411 | — |
| granite-3.1-8b-instruct | 1.72 GiB | 1.811 | — |
| Supra-Router-51M | 0.02 GiB | 3.884 | — |
| Maestro1-9B | 2.17 GiB | 2.129 | — |
| BugTraceAI-Apex-G4-26B-Master-f16 | 7.98 GiB | 2.717 | — |
| Supra-50M-Base | 0.02 GiB | 3.884 | — |
| granite-3.1-2b-instruct | 0.56 GiB | 1.889 | — |
| Qwen3-Coder-Next-Opus-4.6-Reasoning-Distilled | 16.12 GiB | — | — |
●From the filewhat these mean