bunnycore/DeepQwen-3B-LCoT-SCE

Full fine-tune vs bunnycore/QwQen-3B-LCoT. Fine-tune: 100% of the weights changed, across 36 of 36 layers.

Tensors changed: Norms 3/73, MLP 108/108, Attention 145/252

3 derivatives checked

• mradermacher/DeepQwen-3B-LCoT-SCE-GGUF: Same weights, converted
• mradermacher/DeepQwen-3B-LCoT-SCE-i1-GGUF: Quantization
• bunnycore/DeepQwen-3B-LCoT-SCE-R1: Partial fine-tune

On the Hub · Downloads on Model Pulse