We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
The Qwen/Qwen2.5-32B-Instruct model file size is 66g,but the s1-32b is 129g? why? fp32?