Gille/StrangeMerges_50-7B-slerp

Full fine-tune vs Gille/StrangeMerges_49-7B-dare_ties. 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

2 derivatives checked

• mradermacher/StrangeMerges_50-7B-slerp-GGUF: Same weights, converted
• featherless-ai-quants/Gille-StrangeMerges_50-7B-slerp-GGUF: Quantization

On the Hub · Downloads on Model Pulse