littlemex/prismyra-decision-qwen3.6-35b-a3b-nvfp4-36l

Partial fine-tune (fewer layers) vs Qwen/Qwen3.6-35B-A3B-FP8. littlemex/prismyra-decision-qwen3.6-35b-a3b-nvfp4-36l matches part of Qwen/Qwen3.6-35B-A3B-FP8 (2670 of its 64196 tensors; the rest of Qwen/Qwen3.6-35B-A3B-FP8 has no counterpart). On the shared part: Partial change: 29% of the weights changed (attention, mlp), in 36 of 36 layers.

Tensors changed: Norms 0/234, Attention 234/485, MLP 216/463, other 0/1, Embeddings 0/5, lm_head 0/1

On the Hub · Downloads on Model Pulse