metadata
library_name: transformers
tags:
- Uncensored
- Abliterated
- Cubed Reasoning
- QwQ-32B
- reasoning
- thinking
- r1
- cot
- deepseek
- Qwen2.5
- Hermes
- DeepHermes
- DeepSeek
- DeepSeek-R1-Distill
- 128k context
- merge
base_model:
- huihui-ai/QwQ-32B-abliterated
- huihui-ai/DeepSeek-R1-Distill-Qwen-32B-abliterated
- huihui-ai/TinyR1-32B-Preview-abliterated
Qwen2.5-QwQ-35B-Eureka-Cubed-abliterated-uncensored
"Cubed" is an enhanced version of QwQ-32B (Qwen's off the chart reasoning/thinking model) for all use cases.
This is the "abliterated" and "uncensored" version.
NOTE: All 3 models used in this "CUBED" version are "abliterated" and "uncensored".
Special thanks to "huihui-ai" for the hard work and excellent models!
For more information on "CUBED 35B" (normal version, all notes, information applies to this version) version see:
https://huggingface.co/DavidAU/Qwen2.5-QwQ-35B-Eureka-Cubed
and for GGUFS:
https://huggingface.co/DavidAU/Qwen2.5-QwQ-35B-Eureka-Cubed-gguf
Additional model card updates, examples and GGUFS to follow.