Quantization format

UD_TQ1_0

UD_TQ1_0 has a nominal rate of bits per weight. We have no measured files carrying this label yet.

From the file· 50 files measured
Nominal bpw
from the block layout
Measured average
2.067
50 files
Range
1.810–2.771
varies by architecture
File sizes
1.81 GiB+
up to 229.90 GiB

Real files

one per model, most downloaded first
ModelSizeEffective bpwvs nominal
Qwen3-Coder-30B-A3B-Instruct7.46 GiB2.098
Qwen3-VL-30B-A3B-Instruct7.60 GiB2.102
Qwen3-Coder-Next17.64 GiB1.902
GLM-4.7-Flash7.76 GiB2.135
Qwen3-30B-A3B-Instruct-25077.54 GiB2.121
Llama-4-Scout-17B-16E-Instruct27.25 GiB2.155
Kimi-K2-Instruct226.87 GiB1.899
GLM-4.5-Air35.63 GiB2.771
Qwen3-Next-80B-A3B-Instruct19.06 GiB2.013
MiniMax-M2.551.94 GiB1.951
grok-276.17 GiB2.428
GLM-4.7-Flash-REAP-23B-A3B6.09 GiB2.274
DeepSeek-V3.1158.79 GiB1.993
Qwen3-Next-80B-A3B-Thinking18.83 GiB1.989
granite-4.0-h-tiny1.81 GiB2.235
GLM-4.6V34.27 GiB2.733
DeepSeek-V3.2150.20 GiB1.883
Qwen3-VL-30B-A3B-Thinking7.59 GiB2.098
Llama-4-Maverick-17B-128E-Instruct98.45 GiB2.106
Qwen3-VL-235B-A22B-Instruct50.67 GiB1.847
Qwen3-VL-235B-A22B-Thinking50.83 GiB1.853
GLM-4.778.69 GiB1.886
Qwen3-30B-A3B-Thinking-25077.59 GiB2.135
DeepSeek-R1-0528150.52 GiB1.889
ERNIE-4.5-21B-A3B-PT6.06 GiB2.371
From the filewhat these mean

Other quantizations