Gille/StrangeMerges_30-7B-slerp
Full fine-tune vs Gille/StrangeMerges_21-7B-slerp. Fine-tune: 97% of the weights changed, across 32 of 32 layers.
Tensors changed: lm_head 1/1, Norms 65/65, MLP 93/96, Attention 124/128, Embeddings 1/1
3 derivatives checked
- mradermacher/StrangeMerges_30-7B-slerp-GGUF: Same weights, converted
- automerger/Experiment26Strangemerges_30-7B: Full fine-tune
- automerger/Strangemerges_32Strangemerges_30-7B: Full fine-tune