MaziyarPanahi/YamshadowStrangemerges_32_AlloyingotneoyPercival_01
Full fine-tune vs automerger/AlloyingotneoyPercival_01-7B. Fine-tune: 96% of the weights changed, across 32 of 32 layers.
Tensors changed: Embeddings 1/1, Norms 38/65, MLP 92/96, Attention 123/128, lm_head 1/1
1 derivatives checked
- mradermacher/YamshadowStrangemerges_32_AlloyingotneoyPercival_01-GGUF: Same weights, converted
2 other repos hold byte-for-byte identical weight files.