dalatexcoder/Qwen3.5-2B-heretic-ara
Full fine-tune vs Qwen/Qwen3.5-2B-Base. Fine-tune: 100% of the weights changed, across 24 of 24 layers.
Tensors changed: Attention 142/168, Norms 79/79, Embeddings 1/1, MLP 72/72
2 derivatives checked
- mradermacher/Qwen3.5-2B-heretic-ara-i1-GGUF: Quantization
- mradermacher/Qwen3.5-2B-heretic-ara-GGUF: Same weights, converted