ehristoforu/llama-3-12b-instruct
Full fine-tune (fewer layers) vs IlyaGusev/saiga_llama3_8b. IlyaGusev/saiga_llama3_8b matches part of ehristoforu/llama-3-12b-instruct (291 of its 453 tensors; the rest of ehristoforu/llama-3-12b-instruct 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 0/65, MLP 96/96, Attention 128/128, lm_head 1/1
1 derivatives checked
- Letemoin/L3-Deluxe-Smurge_l3-12B: Partial fine-tune (fewer layers)