nineninesix/gepard-1.0
Full fine-tune (fewer layers) vs nineninesix/qwen3_5-full-attn-only-14. nineninesix/qwen3_5-full-attn-only-14 matches part of nineninesix/gepard-1.0 (156 of its 296 tensors; the rest of nineninesix/gepard-1.0 has no counterpart). On the shared part: Fine-tune: 100% of the weights changed, across 14 of 14 layers.
Tensors changed: Embeddings 1/1, Norms 57/57, MLP 42/42, Attention 56/56
2 derivatives checked
- nineninesix/gepard-1.1: LoRA-style
- texdata/Gepard-Slovenian-TTS: Partial fine-tune