inference-optimization/Qwen3.6-8B-A1.6B
Full fine-tune (fewer layers) vs Qwen/Qwen3.6-35B-A3B. inference-optimization/Qwen3.6-8B-A1.6B matches part of Qwen/Qwen3.6-35B-A3B (174 of its 1045 tensors; the rest of Qwen/Qwen3.6-35B-A3B has no counterpart). On the shared part: Fine-tune: 100% of the weights changed, across 8 of 8 layers.
Tensors changed: Embeddings 4/4, Attention 64/64, MLP 68/68, Norms 37/37, lm_head 1/1