bedderautomation/qwen25-3b-abliterated

Partial fine-tune vs Qwen/Qwen2.5-3B-Instruct. Partial change: 33% of the weights changed (attention, embeddings, mlp), in 9 of 36 layers.

Tensors changed: Embeddings 1/1, Norms 0/73, MLP 27/108, Attention 45/252

2 derivatives checked

• mradermacher/qwen25-3b-abliterated-i1-GGUF: Quantization
• mradermacher/qwen25-3b-abliterated-GGUF: Same weights, converted

On the Hub · Downloads on Model Pulse