Qihoo 360's first-generation reasoning model, Tiny-R1-32B-Preview, which outperforms the 70B model Deepseek-R1-Distill-Llama-70B and nearly matches the full R1 model in math.
32b
41 Pulls Updated 5 weeks ago
Updated 5 weeks ago
5 weeks ago
92a69ba150e5 · 20GB
model
archqwen2
·
parameters32.8B
·
quantizationQ4_K_M
20GB
params
{
"repeat_penalty": 1.25,
"stop": [
"<｜end▁of▁sentence｜>",
"<|endoft
130B
template
{{ if .System }}<|im_start|>system
{{ .System }}<|im_end|>
{{ end }}{{ if .Prompt }}<|im_start|>us
186B
Readme
No readme