appvoid/merging-x3
Full fine-tune (fewer layers) vs TinyLlama/TinyLlama-1.1B-intermediate-step-1431k-3T. TinyLlama/TinyLlama-1.1B-intermediate-step-1431k-3T matches part of appvoid/merging-x3 (201 of its 318 tensors; the rest of appvoid/merging-x3 has no counterpart). On the shared part: Fine-tune: 100% of the weights changed, across 22 of 22 layers.
Tensors changed: lm_head 1/1, Embeddings 1/1, Norms 45/45, MLP 66/66, Attention 88/88