cyberagent/CAT-Thinking-8B

Full fine-tune vs Qwen/Qwen3-8B. Fine-tune: 100% of the weights changed, across 36 of 36 layers.

Tensors changed: Embeddings 1/1, Norms 145/145, MLP 108/108, Attention 144/144, lm_head 1/1

3 derivatives checked

• mmnga-o/CAT-Thinking-8B-gguf: Quantization
• mradermacher/CAT-Thinking-8B-i1-GGUF: Quantization
• mradermacher/CAT-Thinking-8B-GGUF: LoRA-style

On the Hub · Downloads on Model Pulse