csuhan/Tar-7B-v0.1
Full fine-tune (fewer layers) vs Qwen/Qwen2.5-7B-Instruct. Qwen/Qwen2.5-7B-Instruct matches part of csuhan/Tar-7B-v0.1 (337 of its 870 tensors; the rest of csuhan/Tar-7B-v0.1 has no counterpart). On the shared part: Fine-tune: 100% of the weights changed, across 28 of 28 layers.
Tensors changed: Norms 57/57, MLP 84/84, Attention 196/196
1 other repos hold byte-for-byte identical weight files.