jwest33/qwen3.5-4b-gabliterated
Partial fine-tune vs Qwen/Qwen3.5-4B. Partial change: 83% of the weights changed (attention, embeddings, mlp), in 28 of 32 layers.
Tensors changed: Embeddings 1/4, MLP 84/196, Attention 133/320, Norms 0/203
3 derivatives checked
- mradermacher/qwen3.5-4b-gabliterated-GGUF: Same weights, converted
- mradermacher/qwen3.5-4b-gabliterated-i1-GGUF: Quantization
- manuojvv/Qwen3.5-4B-gabliterated-Q8: Quantization