s1 is a reasoning model finetuned from Qwen2.5-32B-Instruct on just 1,000 examples. It matches o1-preview & exhibits test-time scaling via budget forcing.
tools
32b
237 Pulls Updated 4 weeks ago
Updated 4 weeks ago
4 weeks ago
a75e926fb690 · 23GB
model
archqwen2
·
parameters32.8B
·
quantizationQ5_K_M
23GB
system
You are Qwen, created by Alibaba Cloud. You are a helpful assistant.
68B
template
{{- if .Messages }}
{{- if or .System .Tools }}<|im_start|>system
{{- if .System }}
{{ .System }}
{{
1.5kB
license
Apache License
Version 2.0, January 200
11kB
Readme
This is an uncensored version of simplescaling/s1-32B created with abliteration (see remove-refusals-with-transformers to know more about it).
This is a crude, proof-of-concept implementation to remove refusals from an LLM model without using TransformerLens.
References
Donation
Your donation helps us continue our further development and improvement, a cup of coffee can do it.
- bitcoin:
bc1qqnkhuchxw0zqjh2ku3lu4hq45hc6gy84uk70ge