Applied-Innovation-Center/Karnak-6B-v1.0
Full fine-tune (fewer layers) vs Qwen/Qwen3-4B-Instruct-2507. Qwen/Qwen3-4B-Instruct-2507 matches part of Applied-Innovation-Center/Karnak-6B-v1.0 (397 of its 596 tensors; the rest of Applied-Innovation-Center/Karnak-6B-v1.0 has no counterpart). On the shared part: Fine-tune: 100% of the weights changed, across 36 of 36 layers.
Tensors changed: Norms 145/145, MLP 108/108, Attention 144/144
3 derivatives checked
- mradermacher/Karnak-6B-v1.0-GGUF: Same weights, converted
- mradermacher/Karnak-6B-v1.0-i1-GGUF: Quantization
- NightPrince/Muslim-6B-PRO: LoRA-style
1 other repos hold byte-for-byte identical weight files.