wjbmattingly/nara-qwen-3.5-2b

Full fine-tune vs Qwen/Qwen3.5-2B. Fine-tune: 100% of the weights changed, across 24 of 24 layers.

Tensors changed: Attention 230/264, Norms 114/177, Embeddings 4/4, MLP 169/172

1 derivatives checked

• mradermacher/nara-qwen-3.5-2b-GGUF: Same weights, converted

On the Hub · Downloads on Model Pulse