nllg/EdiTikZ-9B-RL
Partial fine-tune vs nllg/EdiTikZ-9B. Partial change: 89% of the weights changed (attention, embeddings, lm_head, mlp, norms), in 32 of 32 layers.
Tensors changed: lm_head 1/1, Embeddings 3/4, Norms 213/215, Attention 312/332, MLP 207/208
1 derivatives checked
- mradermacher/EdiTikZ-9B-RL-GGUF: Same weights, converted