nllg/EdiTikZ-4B

Full fine-tune vs Qwen/Qwen3.5-4B. Fine-tune: 100% of the weights changed, across 32 of 32 layers.

Tensors changed: Embeddings 4/4, MLP 196/196, Attention 320/320, Norms 203/203

2 derivatives checked

• mradermacher/EdiTikZ-4B-GGUF: Same weights, converted
• nllg/EdiTikZ-4B-RL: Partial fine-tune

On the Hub · Downloads on Model Pulse