issai/Qwen3.5-4B-Kazakh
Full fine-tune (fewer layers) vs issai/Qwen3.5-4B-Base-Kazakh. issai/Qwen3.5-4B-Base-Kazakh matches part of issai/Qwen3.5-4B-Kazakh (426 of its 723 tensors; the rest of issai/Qwen3.5-4B-Kazakh has no counterpart). On the shared part: Fine-tune: 100% of the weights changed, across 32 of 32 layers.
Tensors changed: Embeddings 1/1, Norms 85/105, Attention 202/224, MLP 96/96
1 derivatives checked
- mradermacher/Qwen3.5-4B-Kazakh-GGUF: Same weights, converted