kilian-group/KBevo-Qwen3-4B-GRPO

Full fine-tune vs kilian-group/KBevo-Qwen3-4B-SFT. Fine-tune: 100% of the weights changed, across 36 of 36 layers.

Tensors changed: Embeddings 1/1, Norms 131/145, MLP 108/108, Attention 144/144

1 derivatives checked

• mradermacher/KBevo-Qwen3-4B-GRPO-GGUF: Same weights, converted

On the Hub · Downloads on Model Pulse